diff --git a/docs/api/base_spider.rst b/docs/api/base_spider.rst
index 8dca7971..510e9de3 100644
--- a/docs/api/base_spider.rst
+++ b/docs/api/base_spider.rst
@@ -4,4 +4,3 @@ Base Spider
 .. automodule:: kingfisher_scrapy.base_spider
    :members:
    :undoc-members:
-
diff --git a/docs/api/exceptions.rst b/docs/api/exceptions.rst
index 788097a2..f360575c 100644
--- a/docs/api/exceptions.rst
+++ b/docs/api/exceptions.rst
@@ -4,4 +4,3 @@ Exceptions
 .. automodule:: kingfisher_scrapy.exceptions
    :members:
    :undoc-members:
-
diff --git a/docs/api/extensions.rst b/docs/api/extensions.rst
new file mode 100644
index 00000000..ab745966
--- /dev/null
+++ b/docs/api/extensions.rst
@@ -0,0 +1,6 @@
+Extensions
+==========
+
+.. automodule:: kingfisher_scrapy.extensions
+   :members:
+   :undoc-members:
diff --git a/docs/api/index.rst b/docs/api/index.rst
index 42f566e7..e7c0c60a 100644
--- a/docs/api/index.rst
+++ b/docs/api/index.rst
@@ -4,4 +4,6 @@ API Reference
 .. toctree::
 
    base_spider.rst
+   extensions.rst
+   util.rst
    exceptions.rst
diff --git a/docs/api/util.rst b/docs/api/util.rst
new file mode 100644
index 00000000..c92cae2c
--- /dev/null
+++ b/docs/api/util.rst
@@ -0,0 +1,6 @@
+Utilities
+=========
+
+.. automodule:: kingfisher_scrapy.util
+   :members:
+   :undoc-members:
diff --git a/docs/writing-spiders.rst b/docs/writing-spiders.rst
index 24aeab18..480fec07 100644
--- a/docs/writing-spiders.rst
+++ b/docs/writing-spiders.rst
@@ -53,18 +53,16 @@ Here is a sample:
 .. code-block:: python
 
     from kingfisher_scrapy.base_spider import SimpleSpider
-    from kingfisher_scrapy.util import handle_error
+    from kingfisher_scrapy.util import components, handle_error
 
     class VerySimple(SimpleSpider):
         name = 'very_simple'
         data_type = 'release_package'
 
         def start_requests(self):
-            # This API only has one URL to get. Make a request for that, and set a filename
-            yield scrapy.Request(
-                'https://buyandsell.gc.ca/cds/public/ocds/tpsgc-pwgsc_ocds_EF-FY-13-14.json',
-                meta={'kf_filename': '13-14.json'}
-            )
+            # Request the source's only URL, and transform the URL to a file name using ``basename``.
+            url = 'https://buyandsell.gc.ca/cds/public/ocds/tpsgc-pwgsc_ocds_EF-FY-13-14.json'
+            yield self.build_request(url, formatter=components(-1))
 
 Spider properties
 -----------------
diff --git a/kingfisher_scrapy/base_spider.py b/kingfisher_scrapy/base_spider.py
index 78163e16..e2db8936 100644
--- a/kingfisher_scrapy/base_spider.py
+++ b/kingfisher_scrapy/base_spider.py
@@ -1,4 +1,3 @@
-import hashlib
 import json
 from datetime import datetime
 from io import BytesIO
@@ -102,6 +101,49 @@ def get_start_time(self, format):
         """
         return self.crawler.stats.get_value('start_time').strftime(format)
 
+    def build_request(self, url, formatter, **kwargs):
+        """
+        Returns a Scrapy request, with a file name added to the request's ``meta`` attribute. If the file name doesn't
+        have a ``.json`` or ``.zip`` extension, it adds a ``.json`` extension.
+
+        If the last component of a URL's path is unique, use it as the file name. For example:
+
+        >>> from kingfisher_scrapy.base_spider import BaseSpider
+        >>> from kingfisher_scrapy.util import components
+        >>> url = 'https://example.com/package.json'
+        >>> formatter = components(-1)
+        >>> BaseSpider(name='my_spider').build_request(url, formatter=formatter).meta
+        {'kf_filename': 'package.json'}
+
+        To use a query string parameter as the file name:
+
+        >>> from kingfisher_scrapy.util import parameters
+        >>> url = 'https://example.com/packages?page=1&per_page=100'
+        >>> formatter = parameters('page')
+        >>> BaseSpider(name='my_spider').build_request(url, formatter=formatter).meta
+        {'kf_filename': 'page-1.json'}
+
+        To add a query string parameter to the file name:
+
+        >>> from kingfisher_scrapy.util import join
+        >>> url = 'https://example.com/packages?page=1&per_page=100'
+        >>> formatter = join(components(-1), parameters('page'))
+        >>> BaseSpider(name='my_spider').build_request(url, formatter=formatter).meta
+        {'kf_filename': 'packages-page-1.json'}
+
+        :param str url: the URL to request
+        :param formatter: a function that accepts a URL and returns a file name
+        :returns: a Scrapy request
+        :rtype: scrapy.Request
+        """
+        file_name = formatter(url)
+        if not file_name.endswith(('.json', '.zip')):
+            file_name += '.json'
+        meta = {'kf_filename': file_name}
+        if 'meta' in kwargs:
+            meta.update(kwargs.pop('meta'))
+        return scrapy.Request(url, meta=meta, **kwargs)
+
     def build_file_from_response(self, response, **kwargs):
         """
         Returns an item to yield, based on the response to a request.
@@ -266,8 +308,7 @@ def start_requests(self):
     @handle_error
     def parse(self, response):
         if self.zip_file_format:
-            filename = '{}.zip'.format(hashlib.md5(response.url.encode('utf-8')).hexdigest())
-            self.build_file_from_response(response, file_name=filename, post_to_api=False)
+            self.build_file_from_response(response, data_type='zip', post_to_api=False)
 
         zip_file = ZipFile(BytesIO(response.body))
         for finfo in zip_file.infolist():
@@ -300,6 +341,8 @@ class LinksSpider(SimpleSpider):
 
     1. Inherit from ``LinksSpider``
     1. Set a ``data_type`` class attribute to the data type of the API responses
+    1. Set a ``next_page_formatter`` class attribute to set the file name as in
+       :meth:`~kingfisher_scrapy.base_spider.BaseSpider.build_request`
     1. Optionally, set a ``next_pointer`` class attribute to the JSON Pointer for the next link (default "/links/next")
     1. Write a ``start_requests`` method to request the first page of API results
 
@@ -333,4 +376,4 @@ def next_link(self, response):
         data = json.loads(response.text)
         url = resolve_pointer(data, self.next_pointer, None)
         if url:
-            return scrapy.Request(url, meta={'kf_filename': hashlib.md5(url.encode('utf-8')).hexdigest() + '.json'})
+            return self.build_request(url, formatter=self.next_page_formatter)
diff --git a/kingfisher_scrapy/spiders/__init__.py b/kingfisher_scrapy/spiders/__init__.py
index ebd689ac..e69de29b 100644
--- a/kingfisher_scrapy/spiders/__init__.py
+++ b/kingfisher_scrapy/spiders/__init__.py
@@ -1,4 +0,0 @@
-# This package will contain the spiders of your Scrapy project
-#
-# Please refer to the documentation for information on how to create and manage
-# your spiders.
diff --git a/kingfisher_scrapy/spiders/afghanistan_records.py b/kingfisher_scrapy/spiders/afghanistan_records.py
index a662dc01..d009b066 100644
--- a/kingfisher_scrapy/spiders/afghanistan_records.py
+++ b/kingfisher_scrapy/spiders/afghanistan_records.py
@@ -3,7 +3,7 @@
 import scrapy
 
 from kingfisher_scrapy.base_spider import SimpleSpider
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import components, handle_error
 
 
 class AfghanistanRecords(SimpleSpider):
@@ -13,17 +13,15 @@ class AfghanistanRecords(SimpleSpider):
     download_delay = 1
 
     def start_requests(self):
-        yield scrapy.Request(
-            'https://ocds.ageops.net/api/ocds/records',
-            meta={'kf_filename': 'list.json'},
-            callback=self.parse_list
-        )
+        # A JSON array of URL strings, in reverse chronological order.
+        url = 'https://ocds.ageops.net/api/ocds/records'
+        yield scrapy.Request(url, meta={'kf_filename': 'list.json'}, callback=self.parse_list)
 
     @handle_error
     def parse_list(self, response):
-        files_urls = json.loads(response.text)
+        urls = json.loads(response.text)
         if self.sample:
-            files_urls = [files_urls[0]]
-
-        for file_url in files_urls:
-            yield scrapy.Request(file_url, meta={'kf_filename': file_url.split('/')[-1] + '.json'})
+            urls = [urls[0]]
+        for url in urls:
+            # URL looks like https://ocds.ageops.net/api/record/5ed2a62c4192f32c8c74a4e5
+            yield self.build_request(url, formatter=components(-1))
diff --git a/kingfisher_scrapy/spiders/afghanistan_releases.py b/kingfisher_scrapy/spiders/afghanistan_releases.py
index 99a49b69..c9755dea 100644
--- a/kingfisher_scrapy/spiders/afghanistan_releases.py
+++ b/kingfisher_scrapy/spiders/afghanistan_releases.py
@@ -3,7 +3,7 @@
 import scrapy
 
 from kingfisher_scrapy.base_spider import SimpleSpider
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import components, handle_error
 
 
 class AfghanistanReleases(SimpleSpider):
@@ -13,30 +13,25 @@ class AfghanistanReleases(SimpleSpider):
     download_delay = 1.5
 
     def start_requests(self):
-        yield scrapy.Request(
-            'https://ocds.ageops.net/api/ocds/releases/dates',
-            meta={'kf_filename': 'list.json'},
-            callback=self.parse_list
-        )
+        # A JSON array of URL strings, in reverse chronological order.
+        url = 'https://ocds.ageops.net/api/ocds/releases/dates'
+        yield scrapy.Request(url, meta={'kf_filename': 'list.json'}, callback=self.parse_list)
 
     @handle_error
     def parse_list(self, response):
-        files_urls = json.loads(response.text)
+        urls = json.loads(response.text)
         if self.sample:
-            files_urls = [files_urls[0]]
-
-        for file_url in files_urls:
-            yield scrapy.Request(
-                file_url,
-                meta={'kf_filename': file_url.split('/')[-1] + '.json'},
-                callback=self.parse_release_list
-            )
+            urls = [urls[0]]
+        for url in urls:
+            # A JSON array of URL strings, in reverse chronological order.
+            # URL looks like https://ocds.ageops.net/api/ocds/releases/2020-05-30
+            yield self.build_request(url, formatter=components(-1), callback=self.parse_release_list)
 
     @handle_error
     def parse_release_list(self, response):
-        files_urls = json.loads(response.text)
+        urls = json.loads(response.text)
         if self.sample:
-            files_urls = [files_urls[0]]
-
-        for file_url in files_urls:
-            yield scrapy.Request(file_url, meta={'kf_filename': file_url.split('/')[-1] + '.json'})
+            urls = [urls[0]]
+        for url in urls:
+            # URL looks like https://ocds.ageops.net/api/release/5ed2a62c4192f32c8c74a4e3
+            yield self.build_request(url, formatter=components(-1))
diff --git a/kingfisher_scrapy/spiders/argentina_buenos_aires.py b/kingfisher_scrapy/spiders/argentina_buenos_aires.py
index 10dadbf5..8ab14456 100644
--- a/kingfisher_scrapy/spiders/argentina_buenos_aires.py
+++ b/kingfisher_scrapy/spiders/argentina_buenos_aires.py
@@ -3,7 +3,7 @@
 import scrapy
 
 from kingfisher_scrapy.base_spider import ZipSpider
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import components, handle_error
 
 
 class ArgentinaBuenosAires(ZipSpider):
@@ -24,15 +24,14 @@ class ArgentinaBuenosAires(ZipSpider):
     download_timeout = 1000
 
     def start_requests(self):
-        yield scrapy.Request(
-            'https://data.buenosaires.gob.ar/api/3/action/package_show?id=buenos-aires-compras',
-            meta={'kf_filename': 'list.json'},
-            callback=self.parse_list
-        )
+        # A CKAN API JSON response.
+        url = 'https://data.buenosaires.gob.ar/api/3/action/package_show?id=buenos-aires-compras'
+        yield scrapy.Request(url, meta={'kf_filename': 'list.json'}, callback=self.parse_list)
 
     @handle_error
     def parse_list(self, response):
         data = json.loads(response.text)
         for resource in data['result']['resources']:
             if resource['format'].upper() == 'JSON':
-                yield scrapy.Request(resource['url'], meta={'kf_filename': resource['url'].rsplit('/', 1)[-1]})
+                # Presently, only one URL matches.
+                yield self.build_request(resource['url'], formatter=components(-1))
diff --git a/kingfisher_scrapy/spiders/argentina_vialidad.py b/kingfisher_scrapy/spiders/argentina_vialidad.py
index 1c44cfc8..f3abfe4e 100644
--- a/kingfisher_scrapy/spiders/argentina_vialidad.py
+++ b/kingfisher_scrapy/spiders/argentina_vialidad.py
@@ -8,7 +8,5 @@ class ArgentinaVialidad(SimpleSpider):
     data_type = 'release_package_list'
 
     def start_requests(self):
-        yield scrapy.Request(
-            'https://datosabiertos.vialidad.gob.ar/api/ocds/package/all',
-            meta={'kf_filename': 'all.json'}
-        )
+        url = 'https://datosabiertos.vialidad.gob.ar/api/ocds/package/all'
+        yield scrapy.Request(url, meta={'kf_filename': 'all.json'})
diff --git a/kingfisher_scrapy/spiders/armenia.py b/kingfisher_scrapy/spiders/armenia.py
index 67c42a6e..04437cd2 100644
--- a/kingfisher_scrapy/spiders/armenia.py
+++ b/kingfisher_scrapy/spiders/armenia.py
@@ -1,12 +1,15 @@
 import scrapy
 
 from kingfisher_scrapy.base_spider import LinksSpider
+from kingfisher_scrapy.util import parameters
 
 
 class Armenia(LinksSpider):
     name = 'armenia'
     data_type = 'release_package'
     next_pointer = '/next_page/uri'
+    next_page_formatter = parameters('offset')
 
     def start_requests(self):
-        yield scrapy.Request('https://armeps.am/ocds/release', meta={'kf_filename': 'page1.json'})
+        url = 'https://armeps.am/ocds/release'
+        yield scrapy.Request(url, meta={'kf_filename': 'offset-0.json'})
diff --git a/kingfisher_scrapy/spiders/australia.py b/kingfisher_scrapy/spiders/australia.py
index a6f1da85..60ec5dcf 100644
--- a/kingfisher_scrapy/spiders/australia.py
+++ b/kingfisher_scrapy/spiders/australia.py
@@ -1,26 +1,18 @@
-import datetime
+from datetime import date
 
 import scrapy
 
 from kingfisher_scrapy.base_spider import LinksSpider
+from kingfisher_scrapy.util import parameters
 
 
 class Australia(LinksSpider):
     name = 'australia'
     data_type = 'release_package'
+    next_page_formatter = parameters('cursor')
 
     def start_requests(self):
-        url_prefix = 'https://api.tenders.gov.au/ocds/findByDates/contractPublished/'
+        url = f'https://api.tenders.gov.au/ocds/findByDates/contractPublished/' \
+              f'2004-01-01T00:00:00Z/{date.today().year}-12-31T23:59:59Z'
 
-        if self.sample:
-            yield scrapy.Request(
-                url_prefix + '2018-01-01T00:00:00Z/2018-12-31T23:59:59Z',
-                meta={'kf_filename': 'year-2018.json'}
-            )
-        else:
-            current_year = datetime.datetime.now().year + 1
-            for year in range(2004, current_year):
-                yield scrapy.Request(
-                    url_prefix + '{}-01-01T00:00:00Z/{}-12-31T23:59:59Z'.format(year, year),
-                    meta={'kf_filename': 'year-{}.json'.format(year)}
-                )
+        yield scrapy.Request(url, meta={'kf_filename': 'start.json'})
diff --git a/kingfisher_scrapy/spiders/australia_nsw.py b/kingfisher_scrapy/spiders/australia_nsw.py
index b09326b4..a32bbbce 100644
--- a/kingfisher_scrapy/spiders/australia_nsw.py
+++ b/kingfisher_scrapy/spiders/australia_nsw.py
@@ -1,10 +1,7 @@
-import hashlib
 import json
 
-import scrapy
-
 from kingfisher_scrapy.base_spider import SimpleSpider
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import handle_error, parameters
 
 
 class AustraliaNSW(SimpleSpider):
@@ -12,54 +9,45 @@ class AustraliaNSW(SimpleSpider):
     data_type = 'release_package'
 
     def start_requests(self):
-        release_types = ['planning', 'tender', 'contract']
-        page_limit = 10 if self.sample else 1000
-        url = 'https://tenders.nsw.gov.au/?event=public.api.{}.search&ResultsPerPage={}'
-        for release_type in release_types:
-            yield scrapy.Request(
-                url.format(release_type, page_limit),
-                meta={
-                    'kf_filename': '{}.json'.format(release_type),
-                    'release_type': release_type,
-                },
+        pattern = 'https://tenders.nsw.gov.au/?event=public.api.{}.search&ResultsPerPage=1000'
+        for release_type in ('planning', 'tender', 'contract'):
+            yield self.build_request(
+                pattern.format(release_type),
+                formatter=parameters('event'),
+                meta={'release_type': release_type},
                 callback=self.parse_list
             )
 
     @handle_error
     def parse_list(self, response):
-        json_data = json.loads(response.text)
+        data = json.loads(response.text)
         release_type = response.request.meta['release_type']
 
-        # More Pages?
-        if 'links' in json_data and isinstance(json_data['links'], dict) and 'next' in json_data['links'] \
-                and not self.sample:
-            yield scrapy.Request(
-                json_data['links']['next'],
-                meta={
-                    'kf_filename': hashlib.md5(json_data['links']['next'].encode('utf-8')).hexdigest() + '.json',
-                    'release_type': release_type,
-                },
+        if 'links' in data and isinstance(data['links'], dict) and 'next' in data['links'] and not self.sample:
+            yield self.build_request(
+                data['links']['next'],
+                formatter=parameters('event', 'startRow'),
+                meta={'release_type': release_type},
                 callback=self.parse_list
             )
 
-        # Data?
-        for release in json_data['releases']:
+        for release in data['releases']:
             if release_type == 'planning':
                 uuid = release['tender']['plannedProcurementUUID']
-                yield scrapy.Request(
-                    'https://tenders.nsw.gov.au/?event=public.api.planning.view&PlannedProcurementUUID=%s' % uuid,
-                    meta={'kf_filename': 'plannning-%s.json' % uuid}
+                yield self.build_request(
+                    'https://tenders.nsw.gov.au/?event=public.api.planning.view&PlannedProcurementUUID=' + uuid,
+                    formatter=parameters('event', 'PlannedProcurementUUID')
                 )
-            if release_type == 'tender':
+            elif release_type == 'tender':
                 uuid = release['tender']['RFTUUID']
-                yield scrapy.Request(
-                    'https://tenders.nsw.gov.au/?event=public.api.tender.view&RFTUUID=%s' % uuid,
-                    meta={'kf_filename': 'tender-%s.json' % uuid}
+                yield self.build_request(
+                    'https://tenders.nsw.gov.au/?event=public.api.tender.view&RFTUUID=' + uuid,
+                    formatter=parameters('event', 'RFTUUID')
                 )
-            if release_type == 'contract':
+            elif release_type == 'contract':
                 for award in release['awards']:
                     uuid = award['CNUUID']
-                    yield scrapy.Request(
-                        'https://tenders.nsw.gov.au/?event=public.api.contract.view&CNUUID=%s' % uuid,
-                        meta={'kf_filename': 'contract-%s.json' % uuid}
+                    yield self.build_request(
+                        'https://tenders.nsw.gov.au/?event=public.api.contract.view&CNUUID=' + uuid,
+                        formatter=parameters('event', 'CNUUID')
                     )
diff --git a/kingfisher_scrapy/spiders/canada_buyandsell.py b/kingfisher_scrapy/spiders/canada_buyandsell.py
index e9d311e2..546e7fcb 100644
--- a/kingfisher_scrapy/spiders/canada_buyandsell.py
+++ b/kingfisher_scrapy/spiders/canada_buyandsell.py
@@ -1,6 +1,5 @@
-import scrapy
-
 from kingfisher_scrapy.base_spider import SimpleSpider
+from kingfisher_scrapy.util import components
 
 
 class CanadaBuyAndSell(SimpleSpider):
@@ -8,21 +7,14 @@ class CanadaBuyAndSell(SimpleSpider):
     data_type = 'release_package'
 
     def start_requests(self):
-        yield scrapy.Request(
+        urls = [
             'https://buyandsell.gc.ca/cds/public/ocds/tpsgc-pwgsc_ocds_EF-FY-13-14.json',
-            meta={'kf_filename': '13-14.json'}
-        )
-        if self.sample:
-            return
-        yield scrapy.Request(
             'https://buyandsell.gc.ca/cds/public/ocds/tpsgc-pwgsc_ocds_EF-FY-14-15.json',
-            meta={'kf_filename': '14-15.json'}
-        )
-        yield scrapy.Request(
             'https://buyandsell.gc.ca/cds/public/ocds/tpsgc-pwgsc_ocds_EF-FY-15-16.json',
-            meta={'kf_filename': '15-16.json'}
-        )
-        yield scrapy.Request(
             'https://buyandsell.gc.ca/cds/public/ocds/tpsgc-pwgsc_ocds_EF-FY-16-17.json',
-            meta={'kf_filename': '16-17.json'}
-        )
+        ]
+        if self.sample:
+            urls = [urls[0]]
+
+        for url in urls:
+            yield self.build_request(url, formatter=components(-1))
diff --git a/kingfisher_scrapy/spiders/canada_montreal.py b/kingfisher_scrapy/spiders/canada_montreal.py
index 64366213..7c9bc11a 100644
--- a/kingfisher_scrapy/spiders/canada_montreal.py
+++ b/kingfisher_scrapy/spiders/canada_montreal.py
@@ -2,32 +2,27 @@
 
 import scrapy
 
-from kingfisher_scrapy.base_spider import BaseSpider
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.base_spider import SimpleSpider
+from kingfisher_scrapy.util import handle_error, parameters, replace_parameter
 
 
-class CanadaMontreal(BaseSpider):
+class CanadaMontreal(SimpleSpider):
     name = 'canada_montreal'
-    page_limit = 10000
+    data_type = 'release_package'
+    step = 10000
 
     def start_requests(self):
-        yield scrapy.Request(
-            'https://ville.montreal.qc.ca/vuesurlescontrats/api/releases.json?limit=%d' % self.page_limit,
-            meta={'kf_filename': 'page0.json'}
-        )
+        url = 'https://ville.montreal.qc.ca/vuesurlescontrats/api/releases.json?limit={step}'.format(step=self.step)
+        yield scrapy.Request(url, meta={'kf_filename': 'offset-0.json'}, callback=self.parse_list)
 
     @handle_error
-    def parse(self, response):
-        # Actual data
-        yield self.build_file_from_response(response, data_type='release_package')
+    def parse_list(self, response):
+        yield from self.parse(response)
 
-        # Load more pages?
-        if not self.sample and response.request.meta['kf_filename'] == 'page0.json':
+        if not self.sample:
             data = json.loads(response.text)
+            offset = data['meta']['pagination']['limit']
             total = data['meta']['count']
-            offset = self.page_limit
-            while offset < total:
-                url = 'https://ville.montreal.qc.ca/vuesurlescontrats/api/releases.json?limit=%d&offset=%d' % \
-                      (self.page_limit, offset)
-                yield scrapy.Request(url, meta={'kf_filename': 'page' + str(offset) + '.json'})
-                offset += self.page_limit
+            for offset in range(offset, total, self.step):
+                url = replace_parameter(response.request.url, 'offset', offset)
+                yield self.build_request(url, formatter=parameters('offset'))
diff --git a/kingfisher_scrapy/spiders/chile_base.py b/kingfisher_scrapy/spiders/chile_base.py
index c0f851f6..1ec7fbe6 100644
--- a/kingfisher_scrapy/spiders/chile_base.py
+++ b/kingfisher_scrapy/spiders/chile_base.py
@@ -1,85 +1,71 @@
-import datetime
 import json
+from datetime import date
 
-import scrapy
+from kingfisher_scrapy.base_spider import SimpleSpider
+from kingfisher_scrapy.util import components, date_range_by_month, handle_error
 
-from kingfisher_scrapy.base_spider import BaseSpider
-from kingfisher_scrapy.util import handle_error
 
-
-class ChileCompraBaseSpider(BaseSpider):
+class ChileCompraBaseSpider(SimpleSpider):
     custom_settings = {
         'DOWNLOAD_FAIL_ON_DATALOSS': False,
     }
     download_timeout = 300
     limit = 100
-    base_list_url = 'https://apis.mercadopublico.cl/OCDS/data/listaA%C3%B1oMes/{}/{:02d}/{}/{}'
-    record_url = 'https://apis.mercadopublico.cl/OCDS/data/record/%s'
-    start_year = 2008
+    base_list_url = 'https://apis.mercadopublico.cl/OCDS/data/listaA%C3%B1oMes/{0.year:d}/{0.month:02d}/{1}/{2}'
 
-    def get_year_month_until(self):
-        until_year = datetime.datetime.now().year + 1
-        until_month = datetime.datetime.now().month
+    def start_requests(self):
+        today = date.today()
         if hasattr(self, 'year'):
-            self.start_year = int(self.year)
-            until_year = self.start_year + 1
-            until_month = 12 if self.start_year != datetime.datetime.now().year else until_month
-        return until_year, until_month
+            year = int(self.year)
+            start = date(year, 1, 1)
+            stop = date(year, 12, 1)
+            if year == today.year:
+                stop = stop.replace(month=today.month)
+        else:
+            start = date(2008, 1, 1)
+            stop = today
 
-    def start_requests(self):
         if self.sample:
-            yield scrapy.Request(
-                self.base_list_url.format(2017, 10, 0, 10),
-                meta={'kf_filename': 'list-2017-10.json', 'year': 2017, 'month': 10},
-            )
-            return
+            start = stop
 
-        until_year, until_month = self.get_year_month_until()
-        for year in range(self.start_year, until_year):
-            for month in range(1, 13):
-                # just scrape until the current month when the until year = current year
-                if (until_year - 1) == year and month > until_month:
-                    break
-                yield scrapy.Request(
-                    self.base_list_url.format(year, month, 0, self.limit),
-                    meta={'kf_filename': 'list-{}-{:02d}.json'.format(year, month), 'year': year, 'month': month},
-                )
+        for d in date_range_by_month(start, stop):
+            yield self.build_request(
+                self.base_list_url.format(d, 0, self.limit),
+                formatter=components(-4, -1),
+                meta={
+                    'year': d.year,
+                    'month': d.month,
+                },
+                callback=self.parse_list
+            )
 
     @handle_error
-    def parse(self, response):
+    def parse_list(self, response):
         data = json.loads(response.text)
-        if 'data' in data:
-            for data_item in data['data']:
-                if self.data_type == 'record_package':
-                    yield scrapy.Request(
-                        self.record_url % data_item['ocid'].replace('ocds-70d2nz-', ''),
-                        meta={'kf_filename': 'data-%s-%s.json' % (data_item['ocid'], self.data_type)}
-                    )
-                else:
-                    # the data comes in this format:
-                    # "data": [
-                    #       {
-                    #        "ocid": "",
-                    #        "urlTender": "..",
-                    #        "urlAward": ".."
-                    #        }
-                    #    ]
-                    for stage in list(data_item.keys()):
-                        if 'url' in stage:
-                            name = stage.replace('url', '')
-                            yield scrapy.Request(
-                                data_item[stage],
-                                meta={'kf_filename': 'data-%s-%s.json' % (data_item['ocid'], name)}
-                            )
-            if 'pagination' in data and (data['pagination']['offset'] + self.limit) < data['pagination']['total']:
-                year = response.request.meta['year']
-                month = response.request.meta['month']
-                offset = data['pagination']['offset']
-                yield scrapy.Request(
-                    self.base_list_url.format(year, month, self.limit + offset, self.limit),
-                    meta={'year': year, 'month': month}
-                )
-        elif 'status' in data and data['status'] != 200:
+        if 'status' in data and data['status'] != 200:
             yield self.build_file_error_from_response(response, errors={'http_code': data['status']})
-        else:
-            yield self.build_file_from_response(response, data_type=self.data_type)
+            return
+
+        for item in data['data']:
+            # An item looks like:
+            #
+            # {
+            #   "ocid": "ocds-70d2nz-2359-2-LE19",
+            #   "urlTender": "https://apis.mercadopublico.cl/OCDS/data/tender/2359-2-LE19",
+            #   "urlAward": "https://apis.mercadopublico.cl/OCDS/data/award/2359-2-LE19",
+            #   "urlPlanning": "https://apis.mercadopublico.cl/OCDS/data/planning/2359-2-LE19"
+            # }
+            yield from self.handle_item(item)
+
+        if 'pagination' in data and (data['pagination']['offset'] + self.limit) < data['pagination']['total']:
+            year = response.request.meta['year']
+            month = response.request.meta['month']
+            offset = data['pagination']['offset']
+            yield self.build_request(
+                self.base_list_url.format(date(year, month, 1), offset + self.limit, self.limit),
+                formatter=components(-4, -1),
+                meta={
+                    'year': year,
+                    'month': month,
+                }
+            )
diff --git a/kingfisher_scrapy/spiders/chile_compra_bulk.py b/kingfisher_scrapy/spiders/chile_compra_bulk.py
index 9b5c18b1..854d627d 100644
--- a/kingfisher_scrapy/spiders/chile_compra_bulk.py
+++ b/kingfisher_scrapy/spiders/chile_compra_bulk.py
@@ -1,8 +1,7 @@
-import hashlib
-
-import scrapy
+from datetime import date
 
 from kingfisher_scrapy.base_spider import ZipSpider
+from kingfisher_scrapy.util import components, date_range_by_month
 
 
 class ChileCompraBulk(ZipSpider):
@@ -16,17 +15,12 @@ class ChileCompraBulk(ZipSpider):
     }
 
     def start_requests(self):
-        url = 'https://ocds.blob.core.windows.net/ocds/{}{}.zip'
+        url = 'https://ocds.blob.core.windows.net/ocds/{0.year:d}{0.month:02d}.zip'
+
+        start = date(2009, 1, 1)
+        stop = date.today().replace(day=1)
         if self.sample:
-            years = ['2017']
-            months = ['02']
-        else:
-            years = range(2017, 2020)
-            months = ['0{}'.format(m) if m < 10 else str(m) for m in range(1, 13)]
+            start = stop
 
-        for year in years:
-            for month in months:
-                yield scrapy.Request(
-                    url.format(year, month),
-                    meta={'kf_filename': hashlib.md5((url).encode('utf-8')).hexdigest()}
-                )
+        for d in date_range_by_month(start, stop):
+            yield self.build_request(url.format(d), formatter=components(-1))
diff --git a/kingfisher_scrapy/spiders/chile_compra_records.py b/kingfisher_scrapy/spiders/chile_compra_records.py
index f7320964..14c5d99c 100644
--- a/kingfisher_scrapy/spiders/chile_compra_records.py
+++ b/kingfisher_scrapy/spiders/chile_compra_records.py
@@ -1,6 +1,11 @@
 from kingfisher_scrapy.spiders.chile_base import ChileCompraBaseSpider
+from kingfisher_scrapy.util import components
 
 
 class ChileCompraRecords(ChileCompraBaseSpider):
     name = 'chile_compra_records'
     data_type = 'record_package'
+
+    def handle_item(self, item):
+        url = 'https://apis.mercadopublico.cl/OCDS/data/record/' + item['ocid'].replace('ocds-70d2nz-', '')
+        yield self.build_request(url, formatter=components(-2))
diff --git a/kingfisher_scrapy/spiders/chile_compra_releases.py b/kingfisher_scrapy/spiders/chile_compra_releases.py
index e1082f83..a3467305 100644
--- a/kingfisher_scrapy/spiders/chile_compra_releases.py
+++ b/kingfisher_scrapy/spiders/chile_compra_releases.py
@@ -1,6 +1,12 @@
 from kingfisher_scrapy.spiders.chile_base import ChileCompraBaseSpider
+from kingfisher_scrapy.util import components
 
 
 class ChileCompraReleases(ChileCompraBaseSpider):
     name = 'chile_compra_releases'
     data_type = 'release_package'
+
+    def handle_item(self, item):
+        for key in item:
+            if key.startswith('url'):
+                yield self.build_request(item[key], formatter=components(-2))
diff --git a/kingfisher_scrapy/spiders/colombia.py b/kingfisher_scrapy/spiders/colombia.py
index 9a76d02d..faedc130 100644
--- a/kingfisher_scrapy/spiders/colombia.py
+++ b/kingfisher_scrapy/spiders/colombia.py
@@ -1,4 +1,3 @@
-import hashlib
 import logging
 import time
 from json import JSONDecodeError
@@ -6,22 +5,29 @@
 import scrapy
 
 from kingfisher_scrapy.base_spider import LinksSpider
+from kingfisher_scrapy.util import parameters
 
 
 class Colombia(LinksSpider):
     name = 'colombia'
-    sleep = 120 * 60
+    next_page_formatter = parameters('page')
 
     def start_requests(self):
         base_url = 'https://apiocds.colombiacompra.gov.co:8443/apiCCE2.0/rest/releases'
         if hasattr(self, 'year'):
-            base_url += '/page/{}'.format(int(self.year))
-        base_url += '?page=%d'
+            base_url += f'/page/{int(self.year)}'
+        base_url += '?page={}'
 
-        start_page = 1
+        page = 1
         if hasattr(self, 'page'):
-            start_page = int(self.page)
-        yield scrapy.Request(base_url % start_page, meta={'kf_filename': 'page{}.json'.format(start_page)})
+            page = int(self.page)
+        yield self.build_request(base_url.format(page), formatter=parameters('page'))
+
+    def retry(self, response, reason):
+        url = response.request.url
+        logging.info(reason.format(url=url, status=response.status))
+        time.sleep(120 * 60)
+        yield scrapy.Request(url, dont_filter=True, meta=response.request.meta)
 
     def parse(self, response):
         # In Colombia, every day at certain hour they run a process in their system that drops the database and make
@@ -30,25 +36,13 @@ def parse(self, response):
         # so eventually the spider will always face the service problems. For that, when the problem occurs, (503
         # status or invalid json) we wait 120 minutes and then continue
         try:
-            if response.status == 503 or response.status == 404:
-                url = response.request.url
-                logging.info('Sleeping due error {} in url {}'.format(response.status, url))
-                time.sleep(self.sleep)
-                yield scrapy.Request(
-                    url,
-                    dont_filter=True,
-                    meta={'kf_filename': hashlib.md5(url.encode('utf-8')).hexdigest() + '.json'}
-                )
-            elif self.is_http_success(response):
+            if self.is_http_success(response):
                 yield self.build_file_from_response(response, data_type='release_package')
                 if not self.sample:
                     yield self.next_link(response)
+            elif response.status == 503 or response.status == 404:
+                self.retry(response, 'Sleeping due to HTTP error {status} from {url}')
             else:
                 yield self.build_file_error_from_response(response)
-
         except JSONDecodeError:
-            url = response.request.url
-            logging.info('Sleeping due json decode error in url {}'.format(url))
-            time.sleep(self.sleep)
-            yield scrapy.Request(url, dont_filter=True,
-                                 meta={'kf_filename': hashlib.md5(url.encode('utf-8')).hexdigest() + '.json'})
+            self.retry(response, 'Sleeping due to JSONDecodeError from {url}')
diff --git a/kingfisher_scrapy/spiders/colombia_bulk.py b/kingfisher_scrapy/spiders/colombia_bulk.py
index 2523f8bc..2f1c4036 100644
--- a/kingfisher_scrapy/spiders/colombia_bulk.py
+++ b/kingfisher_scrapy/spiders/colombia_bulk.py
@@ -1,9 +1,7 @@
-from urllib.parse import urlparse
-
 import scrapy
 
 from kingfisher_scrapy.base_spider import ZipSpider
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import components, handle_error
 
 
 class ColombiaBulk(ZipSpider):
@@ -34,8 +32,9 @@ def start_requests(self):
 
     @handle_error
     def parse_list(self, response):
-        urls = response.css('.enlaces_contenido').css('a::attr(href)').getall()
-        urls = [urls[0]] if self.sample else urls
+        urls = response.xpath('//a[@class="enlaces_contenido"]/@href').getall()
+        if self.sample:
+            urls = [urls[0]]
         for url in urls:
-            filename = urlparse(url).path.split('/')[-1]
-            yield scrapy.Request(url, meta={'kf_filename': filename})
+            # URL looks like https://apiocds.colombiacompra.gov.co:8443/ArchivosSECOP/Archivos/SI2011.zip
+            yield self.build_request(url, formatter=components(-1))
diff --git a/kingfisher_scrapy/spiders/digiwhist_base.py b/kingfisher_scrapy/spiders/digiwhist_base.py
index f6d117c0..28a903b6 100644
--- a/kingfisher_scrapy/spiders/digiwhist_base.py
+++ b/kingfisher_scrapy/spiders/digiwhist_base.py
@@ -15,7 +15,7 @@ def start_requests(self):
 
     @handle_error
     def parse(self, response):
-        yield self.build_file_from_response(response, post_to_api=False)
+        yield self.build_file_from_response(response, data_type='tar.gz', post_to_api=False)
 
         # Load a line at the time, pass it to API
         with tarfile.open(fileobj=BytesIO(response.body), mode="r:gz") as tar:
diff --git a/kingfisher_scrapy/spiders/dominican_republic.py b/kingfisher_scrapy/spiders/dominican_republic.py
index 4fa46c4d..18098a6c 100644
--- a/kingfisher_scrapy/spiders/dominican_republic.py
+++ b/kingfisher_scrapy/spiders/dominican_republic.py
@@ -5,7 +5,7 @@
 import scrapy
 
 from kingfisher_scrapy.base_spider import BaseSpider
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import components, handle_error
 
 
 class DominicanRepublic(BaseSpider):
@@ -30,7 +30,7 @@ def parse_list(self, response):
 
         for url in json_urls:
             if '/JSON_DGCP_' in url:
-                yield scrapy.Request('https:' + url, meta={'kf_filename': url.rsplit('/', 1)[-1]})
+                yield self.build_request('https:' + url, formatter=components(-1))
 
     @handle_error
     def parse(self, response):
diff --git a/kingfisher_scrapy/spiders/france.py b/kingfisher_scrapy/spiders/france.py
index a624bffe..21c6fbf6 100644
--- a/kingfisher_scrapy/spiders/france.py
+++ b/kingfisher_scrapy/spiders/france.py
@@ -1,10 +1,9 @@
-import hashlib
 import json
 
 import scrapy
 
 from kingfisher_scrapy.base_spider import SimpleSpider
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import components, handle_error, parameters
 
 
 class France(SimpleSpider):
@@ -12,36 +11,24 @@ class France(SimpleSpider):
     data_type = 'release_package'
 
     def start_requests(self):
-        yield scrapy.Request(
-            'https://www.data.gouv.fr/api/1/datasets/?organization=534fff75a3a7292c64a77de4',
-            meta={'kf_filename': 'list.json'},
-            callback=self.parse_list,
-        )
+        # A CKAN API JSON response.
+        url = 'https://www.data.gouv.fr/api/1/datasets/?organization=534fff75a3a7292c64a77de4'
+        yield scrapy.Request(url, meta={'kf_filename': 'page-1.json'}, callback=self.parse_list)
 
     @handle_error
     def parse_list(self, response):
-        json_data = json.loads(response.text)
-        data = json_data['data']
-        for item in data:
-            resources = item['resources']
-            for resource in resources:
+        data = json.loads(response.text)
+        for item in data['data']:
+            for resource in item['resources']:
                 description = resource['description']
-                if description and (description.count("OCDS") or description.count("ocds")):
-                    url = resource['url']
-                    yield scrapy.Request(
-                        url,
-                        meta={'kf_filename': hashlib.md5(url.encode('utf-8')).hexdigest() + '.json'},
-                    )
+                if description and 'ocds' in description.lower():
+                    yield self.build_request(resource['url'], formatter=components(-2))
                     if self.sample:
                         break
             else:
                 continue
             break
         else:
-            next_page = json_data.get('next_page')
+            next_page = data.get('next_page')
             if next_page:
-                yield scrapy.Request(
-                    next_page,
-                    meta={'kf_filename': hashlib.md5(next_page.encode('utf-8')).hexdigest() + '.json'},
-                    callback=self.parse_list
-                )
+                yield self.build_request(next_page, formatter=parameters('page'), callback=self.parse_list)
diff --git a/kingfisher_scrapy/spiders/georgia_records.py b/kingfisher_scrapy/spiders/georgia_records.py
index 60438bbc..f9f7cef3 100644
--- a/kingfisher_scrapy/spiders/georgia_records.py
+++ b/kingfisher_scrapy/spiders/georgia_records.py
@@ -1,11 +1,14 @@
 import scrapy
 
 from kingfisher_scrapy.base_spider import LinksSpider
+from kingfisher_scrapy.util import parameters
 
 
 class GeorgiaRecords(LinksSpider):
     name = 'georgia_records'
     data_type = 'record_package'
+    next_page_formatter = parameters('page')
 
     def start_requests(self):
-        yield scrapy.Request('https://odapi.spa.ge/api/records.json', meta={'kf_filename': 'page1.json'})
+        url = 'https://odapi.spa.ge/api/records.json'
+        yield scrapy.Request(url, meta={'kf_filename': 'page-1.json'})
diff --git a/kingfisher_scrapy/spiders/georgia_releases.py b/kingfisher_scrapy/spiders/georgia_releases.py
index 6cf0263d..ee3ce55b 100644
--- a/kingfisher_scrapy/spiders/georgia_releases.py
+++ b/kingfisher_scrapy/spiders/georgia_releases.py
@@ -1,11 +1,14 @@
 import scrapy
 
 from kingfisher_scrapy.base_spider import LinksSpider
+from kingfisher_scrapy.util import parameters
 
 
 class GeorgiaReleases(LinksSpider):
     name = 'georgia_releases'
     data_type = 'release_package'
+    next_page_formatter = parameters('page')
 
     def start_requests(self):
-        yield scrapy.Request('https://odapi.spa.ge/api/releases.json', meta={'kf_filename': 'page1.json'})
+        url = 'https://odapi.spa.ge/api/releases.json'
+        yield scrapy.Request(url, meta={'kf_filename': 'page-1.json'})
diff --git a/kingfisher_scrapy/spiders/honduras_cost.py b/kingfisher_scrapy/spiders/honduras_cost.py
index a88e3e4b..329f0273 100644
--- a/kingfisher_scrapy/spiders/honduras_cost.py
+++ b/kingfisher_scrapy/spiders/honduras_cost.py
@@ -1,9 +1,6 @@
-import hashlib
-
 import scrapy
 
 from kingfisher_scrapy.base_spider import SimpleSpider
-from kingfisher_scrapy.util import handle_error
 
 
 class HondurasCoST(SimpleSpider):
@@ -11,22 +8,5 @@ class HondurasCoST(SimpleSpider):
     data_type = 'record_package'
 
     def start_requests(self):
-        yield scrapy.Request(
-            'http://app.sisocs.org/protected/ocdsShow/',
-            meta={'kf_filename': 'list.html'},
-            callback=self.parse_list
-        )
-
-    @handle_error
-    def parse_list(self, response):
-        btns = response.css('script').xpath('text()').getall()
-        for btn in btns:
-            if 'download-all' and 'url:' in btn:
-                array_url = btn.split()
-                for url in array_url:
-                    if 'url:' in url and '?' not in url:
-                        url = url.replace('"', '').replace(',', '').lstrip('url:')
-                        yield scrapy.Request(
-                            url,
-                            meta={'kf_filename': hashlib.md5(url.encode('utf-8')).hexdigest() + '.json'}
-                        )
+        # Extracted from http://app.sisocs.org/protected/ocdsShow/
+        yield scrapy.Request('http://67.207.88.38:8080/sisocs/records', meta={'kf_filename': 'all.json'})
diff --git a/kingfisher_scrapy/spiders/honduras_oncae.py b/kingfisher_scrapy/spiders/honduras_oncae.py
index 91e2347f..611b0421 100644
--- a/kingfisher_scrapy/spiders/honduras_oncae.py
+++ b/kingfisher_scrapy/spiders/honduras_oncae.py
@@ -1,9 +1,7 @@
-from urllib.parse import urlparse
-
 import scrapy
 
 from kingfisher_scrapy.base_spider import ZipSpider
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import components, handle_error
 
 
 class HondurasONCAE(ZipSpider):
@@ -22,11 +20,9 @@ def start_requests(self):
 
     @handle_error
     def parse_list(self, response):
-        urls = response.css(".article-content ul")\
-            .xpath(".//a[contains(., '[json]')]/@href")\
-            .getall()
+        urls = response.xpath('//a[contains(., "[json]")]/@href').getall()
         if self.sample:
             urls = [urls[0]]
         for url in urls:
-            filename = urlparse(url).path.split('/')[-1]
-            yield scrapy.Request(url, meta={'kf_filename': filename})
+            # URL looks like http://200.13.162.79/datosabiertos/HC1/HC1_datos_2020_json.zip
+            yield self.build_request(url, formatter=components(-1))
diff --git a/kingfisher_scrapy/spiders/honduras_portal_bulk_files.py b/kingfisher_scrapy/spiders/honduras_portal_bulk_files.py
index 55195d82..f0e1f9f3 100644
--- a/kingfisher_scrapy/spiders/honduras_portal_bulk_files.py
+++ b/kingfisher_scrapy/spiders/honduras_portal_bulk_files.py
@@ -3,7 +3,7 @@
 import scrapy
 
 from kingfisher_scrapy.base_spider import SimpleSpider
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import components, handle_error
 
 
 class HondurasPortalBulkFiles(SimpleSpider):
@@ -19,13 +19,10 @@ def start_requests(self):
 
     @handle_error
     def parse_list(self, response):
-        filelist = json.loads(response.text)
-
+        items = json.loads(response.text)
         if self.sample:
-            url = filelist[0]['urls']['json']
-            yield scrapy.Request(url, meta={'kf_filename': url.rsplit('/', 1)[-1]})
+            items = [items[0]]
 
-        else:
-            for item in filelist:
-                url = item['urls']['json']
-                yield scrapy.Request(url, meta={'kf_filename': url.rsplit('/', 1)[-1]})
+        for item in items:
+            url = item['urls']['json']
+            yield self.build_request(url, formatter=components(-1))
diff --git a/kingfisher_scrapy/spiders/honduras_portal_records.py b/kingfisher_scrapy/spiders/honduras_portal_records.py
index 1d3fc5de..c5f5ec69 100644
--- a/kingfisher_scrapy/spiders/honduras_portal_records.py
+++ b/kingfisher_scrapy/spiders/honduras_portal_records.py
@@ -1,8 +1,7 @@
-import hashlib
-
 import scrapy
 
 from kingfisher_scrapy.base_spider import LinksSpider
+from kingfisher_scrapy.util import parameters
 
 
 class HondurasPortalRecords(LinksSpider):
@@ -10,9 +9,10 @@ class HondurasPortalRecords(LinksSpider):
     data_type = 'record_package'
     data_pointer = '/recordPackage'
     next_pointer = '/next'
+    next_page_formatter = parameters('page')
 
     download_delay = 0.9
 
     def start_requests(self):
         url = 'http://www.contratacionesabiertas.gob.hn/api/v1/record/?format=json'
-        yield scrapy.Request(url, meta={'kf_filename': hashlib.md5(url.encode('utf-8')).hexdigest() + '.json'})
+        yield scrapy.Request(url, meta={'kf_filename': 'page-1.json'})
diff --git a/kingfisher_scrapy/spiders/honduras_portal_releases.py b/kingfisher_scrapy/spiders/honduras_portal_releases.py
index a676383b..ca4c56f1 100644
--- a/kingfisher_scrapy/spiders/honduras_portal_releases.py
+++ b/kingfisher_scrapy/spiders/honduras_portal_releases.py
@@ -1,8 +1,7 @@
-import hashlib
-
 import scrapy
 
 from kingfisher_scrapy.base_spider import LinksSpider
+from kingfisher_scrapy.util import parameters
 
 
 class HondurasPortalReleases(LinksSpider):
@@ -10,9 +9,10 @@ class HondurasPortalReleases(LinksSpider):
     data_type = 'release_package'
     data_pointer = '/releasePackage'
     next_pointer = '/next'
+    next_page_formatter = parameters('page')
 
     download_delay = 0.9
 
     def start_requests(self):
         url = 'http://www.contratacionesabiertas.gob.hn/api/v1/release/?format=json'
-        yield scrapy.Request(url, meta={'kf_filename': hashlib.md5(url.encode('utf-8')).hexdigest() + '.json'})
+        yield scrapy.Request(url, meta={'kf_filename': 'page-1.json'})
diff --git a/kingfisher_scrapy/spiders/indonesia_bandung.py b/kingfisher_scrapy/spiders/indonesia_bandung.py
index b5af803b..9e1651bd 100644
--- a/kingfisher_scrapy/spiders/indonesia_bandung.py
+++ b/kingfisher_scrapy/spiders/indonesia_bandung.py
@@ -1,51 +1,40 @@
-import datetime
-import hashlib
 import json
-
-import scrapy
+from datetime import date
 
 from kingfisher_scrapy.base_spider import BaseSpider
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import components, date_range_by_year, handle_error, join, parameters
 
 
 class IndonesiaBandung(BaseSpider):
     name = 'indonesia_bandung'
 
     def start_requests(self):
-        url = 'https://birms.bandung.go.id/api/packages/year/{}'
-        current_year = datetime.datetime.now().year + 1
-        for year in range(2013, current_year):
-            yield scrapy.Request(
-                url.format(year),
-                meta={'kf_filename': 'start_requests'},
-                callback=self.parse_data
-            )
+        pattern = 'https://birms.bandung.go.id/api/packages/year/{}'
+
+        start = 2013
+        stop = date.today().year
+
+        for year in date_range_by_year(start, stop):
+            yield self.build_request(pattern.format(year), formatter=components(-1), callback=self.parse_list)
 
     @handle_error
-    def parse_data(self, response):
-        json_data = json.loads(response.text)
-        items = json_data['data']
-        for data in items:
-            url = data['uri']
+    def parse_list(self, response):
+        data = json.loads(response.text)
+        for item in data['data']:
+            url = item['uri']
             if url:
-                yield scrapy.Request(
-                    url,
-                    meta={'kf_filename': hashlib.md5(url.encode('utf-8')).hexdigest() + '.json'},
-                )
+                yield self.build_request(url, formatter=components(-1))
                 if self.sample:
                     break
         else:
-            next_page_url = json_data.get('next_page_url')
+            next_page_url = data.get('next_page_url')
             if next_page_url:
-                yield scrapy.Request(
-                    next_page_url,
-                    meta={'kf_filename': next_page_url.rsplit('/', 1)[-1] + '.json'},
-                    callback=self.parse_data
-                )
+                yield self.build_request(next_page_url, formatter=join(components(-1), parameters('page')),
+                                         callback=self.parse_list)
 
     @handle_error
     def parse(self, response):
-        json_data = json.loads(response.text)
-        if len(json_data) == 0:
+        data = json.loads(response.text)
+        if len(data) == 0:
             return
         yield self.build_file_from_response(response, data_type='release')
diff --git a/kingfisher_scrapy/spiders/kenya_makueni.py b/kingfisher_scrapy/spiders/kenya_makueni.py
index 11d4d9ac..036f7db0 100644
--- a/kingfisher_scrapy/spiders/kenya_makueni.py
+++ b/kingfisher_scrapy/spiders/kenya_makueni.py
@@ -1,39 +1,32 @@
-import hashlib
 from math import ceil
 
 import scrapy
 
 from kingfisher_scrapy.base_spider import SimpleSpider
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import handle_error, parameters
 
 
 class KenyaMakueni(SimpleSpider):
     name = 'kenya_makueni'
     data_type = 'release_package_list'
-    url = 'https://opencontracting.makueni.go.ke/api/ocds/package/all?pageSize={}&pageNumber={}'
+    step = 10
+
+    url = 'https://opencontracting.makueni.go.ke/api/ocds/package/all?pageSize={step}&pageNumber={page}'
 
     def start_requests(self):
         if self.sample:
-            url = self.url.format(10, 0)
-            yield scrapy.Request(
-                url,
-                meta={'kf_filename': hashlib.md5(url.encode('utf-8')).hexdigest() + '.json'}
-            )
+            url = self.url.format(step=self.step, page=0)
+            yield self.build_request(url, formatter=parameters('pageNumber'))
         else:
             yield scrapy.Request(
                 'https://opencontracting.makueni.go.ke/api/ocds/release/count',
-                meta={'kf_filename': 'start_requests'},
+                meta={'kf_filename': 'count.json'},
                 callback=self.parse_count
             )
 
     @handle_error
     def parse_count(self, response):
         total = int(response.text)
-        page_size = 300
-
-        for page_number in range((ceil(total / page_size))):
-            url = self.url.format(page_size, page_number)
-            yield scrapy.Request(
-                url,
-                meta={'kf_filename': hashlib.md5(url.encode('utf-8')).hexdigest() + '.json'}
-            )
+        for page in range(ceil(total / self.step)):
+            url = self.url.format(step=self.step, page=page)
+            yield self.build_request(url, formatter=parameters('pageNumber'))
diff --git a/kingfisher_scrapy/spiders/malta.py b/kingfisher_scrapy/spiders/malta.py
index 1c3dfe4b..70b32400 100644
--- a/kingfisher_scrapy/spiders/malta.py
+++ b/kingfisher_scrapy/spiders/malta.py
@@ -1,11 +1,10 @@
-import hashlib
 import json
-from urllib.parse import urlparse
+from urllib.parse import urlsplit
 
 import scrapy
 
 from kingfisher_scrapy.base_spider import ZipSpider
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import components, handle_error
 
 
 class Malta(ZipSpider):
@@ -15,22 +14,17 @@ class Malta(ZipSpider):
     def start_requests(self):
         yield scrapy.Request(
             'http://demowww.etenders.gov.mt/ocds/services/recordpackage/getrecordpackagelist',
-            meta={'kf_filename': 'start_requests'},
+            meta={'kf_filename': 'list.json'},
             callback=self.parse_list
         )
 
     @handle_error
     def parse_list(self, response):
-        url = 'http://demowww.etenders.gov.mt{}'
-        json_data = json.loads(response.text)
-        packages = json_data['packagesPerMonth']
-        for package in packages:
-            parsed = urlparse(package)
-            path = parsed.path
-            if path:
-                yield scrapy.Request(
-                    url.format(path),
-                    meta={'kf_filename': hashlib.md5(path.encode('utf-8')).hexdigest() + '.json'}
-                )
-                if self.sample:
-                    break
+        urls = json.loads(response.text)['packagesPerMonth']
+        if self.sample:
+            urls = [urls[0]]
+
+        netloc = urlsplit(response.request.url).netloc
+        for url in urls:
+            # URL looks like http://malta-demo-server.eurodyn.com/ocds/services/recordpackage/getrecordpackage/2020/1
+            yield self.build_request(urlsplit(url)._replace(netloc=netloc).geturl(), formatter=components(-2))
diff --git a/kingfisher_scrapy/spiders/mexico_administracion_publica_federal.py b/kingfisher_scrapy/spiders/mexico_administracion_publica_federal.py
index 8d1b98f8..64425413 100644
--- a/kingfisher_scrapy/spiders/mexico_administracion_publica_federal.py
+++ b/kingfisher_scrapy/spiders/mexico_administracion_publica_federal.py
@@ -1,38 +1,32 @@
 import json
+from math import ceil
 
 import scrapy
 
-from kingfisher_scrapy.base_spider import BaseSpider
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.base_spider import SimpleSpider
+from kingfisher_scrapy.util import handle_error, parameters, replace_parameter
 
 
-class MexicoAdministracionPublicaFederal(BaseSpider):
+class MexicoAdministracionPublicaFederal(SimpleSpider):
     """
     Bulk downloads: https://datos.gob.mx/busca/dataset/concentrado-de-contrataciones-abiertas-de-la-apf
     """
     name = 'mexico_administracion_publica_federal'
+    data_type = 'record_package_list_in_results'
 
     def start_requests(self):
-        yield scrapy.Request(
-            'https://api.datos.gob.mx/v1/contratacionesabiertas',
-            meta={'kf_filename': 'page1.json'}
-        )
+        url = 'https://api.datos.gob.mx/v1/contratacionesabiertas'
+        yield scrapy.Request(url, meta={'kf_filename': 'page-1.json'}, callback=self.parse_list)
 
     @handle_error
-    def parse(self, response):
-        data = json.loads(response.text)
+    def parse_list(self, response):
+        yield from self.parse(response)
 
-        # Actual data
-        yield self.build_file_from_response(response, data_type='record_package_list_in_results')
-
-        # Load more pages?
-        if data['pagination']['page'] == 1 and not self.sample:
+        if not self.sample:
+            data = json.loads(response.text)
+            page = data['pagination']['page']
             total = data['pagination']['total']
-            page = 1
             limit = data['pagination']['pageSize']
-            while ((page - 1) * limit) < total:
-                yield scrapy.Request(
-                    'https://api.datos.gob.mx/v1/contratacionesabiertas?page=%d' % page,
-                    meta={'kf_filename': 'page' + str(page) + '.json'}
-                )
-                page += 1
+            for page in range(page + 1, ceil(total / limit)):
+                url = replace_parameter(response.request.url, 'page', page)
+                yield self.build_request(url, formatter=parameters('page'))
diff --git a/kingfisher_scrapy/spiders/mexico_cdmx.py b/kingfisher_scrapy/spiders/mexico_cdmx.py
index 7a96ae64..363e6c64 100644
--- a/kingfisher_scrapy/spiders/mexico_cdmx.py
+++ b/kingfisher_scrapy/spiders/mexico_cdmx.py
@@ -3,7 +3,7 @@
 import scrapy
 
 from kingfisher_scrapy.base_spider import SimpleSpider
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import components, handle_error
 
 
 class MexicoCDMXSource(SimpleSpider):
@@ -19,9 +19,10 @@ def start_requests(self):
 
     @handle_error
     def parse_list(self, response):
-        data = json.loads(response.text)
+        items = json.loads(response.text)
         if self.sample:
-            data = [data[0]]
+            items = [items[0]]
 
-        for data_item in data:
-            yield scrapy.Request(data_item['uri'], meta={'kf_filename': 'id%s.json' % data_item['id']})
+        for item in items:
+            # URL looks like http://www.contratosabiertos.cdmx.gob.mx/api/contrato/OCDS-87SD3T-SEDEMA-LP-0027-2017
+            yield self.build_request(item['uri'], formatter=components(-1))
diff --git a/kingfisher_scrapy/spiders/mexico_grupo_aeroporto.py b/kingfisher_scrapy/spiders/mexico_grupo_aeroporto.py
index 08288d21..5d4ba879 100644
--- a/kingfisher_scrapy/spiders/mexico_grupo_aeroporto.py
+++ b/kingfisher_scrapy/spiders/mexico_grupo_aeroporto.py
@@ -8,7 +8,5 @@ class MexicoGrupoAeroporto(SimpleSpider):
     data_type = 'release_package'
 
     def start_requests(self):
-        yield scrapy.Request(
-            'http://gacmda.gacm.mx:8880/files/opendata/coleccion/concentrado05032019RELEASE.json',
-            meta={'kf_filename': 'concentrado05032019RELEASE.json'}
-        )
+        url = 'http://gacmda.gacm.mx:8880/files/opendata/coleccion/concentrado05032019RELEASE.json'
+        yield scrapy.Request(url, meta={'kf_filename': 'all.json'})
diff --git a/kingfisher_scrapy/spiders/mexico_inai.py b/kingfisher_scrapy/spiders/mexico_inai.py
index 3cb7bbd5..a58cbb09 100644
--- a/kingfisher_scrapy/spiders/mexico_inai.py
+++ b/kingfisher_scrapy/spiders/mexico_inai.py
@@ -1,10 +1,9 @@
-import hashlib
 import json
 
 import scrapy
 
 from kingfisher_scrapy.base_spider import SimpleSpider
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import components, handle_error, parameters
 
 
 class MexicoINAI(SimpleSpider):
@@ -13,6 +12,7 @@ class MexicoINAI(SimpleSpider):
     encoding = 'utf-8-sig'
 
     def start_requests(self):
+        # A CKAN API JSON response.
         yield scrapy.Request(
             'https://datos.gob.mx/busca/api/3/action/package_search?q=organization:inai&rows=500',
             meta={'kf_filename': 'list.json'},
@@ -25,22 +25,12 @@ def parse_list(self, response):
         for result in datas['result']['results']:
             for resource in result['resources']:
                 if resource['format'] == 'JSON':
-                    kf_filename = 'redirect-' + hashlib.md5(resource['url'].encode('utf-8')).hexdigest() + '.json'
-                    yield scrapy.Request(
-                        resource['url'],
-                        meta={
-                            'kf_filename': kf_filename,
-                            'dont_redirect': True
-                        },
-                        callback=self.parse_redirect
-                    )
+                    yield self.build_request(resource['url'], formatter=components(-1), meta={'dont_redirect': True},
+                                             callback=self.parse_redirect)
 
     def parse_redirect(self, response):
         if response.status == 301:
-            url = response.headers['Location'].decode("utf-8").replace("open?", "uc?export=download&")
-            yield scrapy.Request(
-                url,
-                meta={'kf_filename': 'data-' + hashlib.md5(url.encode('utf-8')).hexdigest() + '.json'}
-            )
+            url = response.headers['Location'].decode('utf-8').replace('open?', 'uc?export=download&')
+            yield self.build_request(url, formatter=parameters('id'))
         else:
             yield self.build_file_error_from_response(response)
diff --git a/kingfisher_scrapy/spiders/mexico_jalisco.py b/kingfisher_scrapy/spiders/mexico_jalisco.py
index 7aef33eb..07485b43 100644
--- a/kingfisher_scrapy/spiders/mexico_jalisco.py
+++ b/kingfisher_scrapy/spiders/mexico_jalisco.py
@@ -1,10 +1,9 @@
-import hashlib
 import json
 
 import scrapy
 
 from kingfisher_scrapy.base_spider import SimpleSpider
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import components, handle_error
 
 
 class MexicoJalisco(SimpleSpider):
@@ -20,23 +19,22 @@ def start_requests(self):
 
     @handle_error
     def parse_list(self, response):
-        datas = json.loads(response.text)
+        items = json.loads(response.text)
         if self.sample:
-            datas = [datas[0]]
-        for data in datas:
+            items = [items[0]]
+
+        for item in items:
             yield scrapy.Request(
-                data['URIContract'],
-                meta={'kf_filename': 'id%s.json' % data['ocid']},
+                item['URIContract'],
+                meta={'kf_filename': f"id{item['ocid']}.json"},
                 callback=self.parse_record_package
             )
 
     @handle_error
     def parse_record_package(self, response):
-        json_data = json.loads(response.text)
-        if 'packages' in json_data:
-            for url in json_data['packages']:
-                yield scrapy.Request(
-                    url,
-                    meta={'kf_filename': 'packages-%s.json' % hashlib.md5(url.encode('utf-8')).hexdigest()}
-                )
         yield self.build_file_from_response(response, data_type='record_package')
+
+        data = json.loads(response.text)
+        if 'packages' in data:
+            for url in data['packages']:
+                yield self.build_request(url, formatter=components(-1))
diff --git a/kingfisher_scrapy/spiders/mexico_quien_es_quien.py b/kingfisher_scrapy/spiders/mexico_quien_es_quien.py
index c999d767..920537e2 100644
--- a/kingfisher_scrapy/spiders/mexico_quien_es_quien.py
+++ b/kingfisher_scrapy/spiders/mexico_quien_es_quien.py
@@ -1,52 +1,40 @@
-import hashlib
 import json
 from math import ceil
 
 import scrapy
 
 from kingfisher_scrapy.base_spider import BaseSpider
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import handle_error, parameters
 
 
 class MexicoQuienEsQuien(BaseSpider):
     name = 'mexico_quien_es_quien'
     download_delay = 0.9
-    url = 'https://api.quienesquien.wiki/v2/contracts?limit={}&offset={}'
 
     def start_requests(self):
-        if self.sample:
-            limit = 10
-            offset = 0
-            yield scrapy.Request(
-                self.url.format(limit, offset),
-                meta={'kf_filename': 'sample.json'}
-            )
-        else:
-            yield scrapy.Request(
-                'https://api.quienesquien.wiki/v2/sources',
-                meta={'kf_filename': 'start_requests'},
-                callback=self.parse_count
-            )
+        yield scrapy.Request(
+            'https://api.quienesquien.wiki/v2/sources',
+            meta={'kf_filename': 'list.json'},
+            callback=self.parse_list
+        )
 
     @handle_error
-    def parse_count(self, response):
+    def parse_list(self, response):
+        pattern = 'https://api.quienesquien.wiki/v2/contracts?limit={limit}&offset={offset}'
         limit = 1000
-        json_data = json.loads(response.text)
-        count_list = json_data['data']
-        count = int(count_list[0]['collections']['contracts']['count'])
 
+        count = json.loads(response.text)['data'][0]['collections']['contracts']['count']
         for offset in range(ceil(count / limit)):
-            yield scrapy.Request(
-                self.url.format(limit, (offset * limit)),
-                meta={'kf_filename': hashlib.md5((self.url +
-                                                  str(offset)).encode('utf-8')).hexdigest() + '.json'}
-            )
+            url = pattern.format(limit=limit, offset=offset * limit)
+            yield self.build_request(url, formatter=parameters('offset'))
+            if self.sample:
+                break
 
     @handle_error
     def parse(self, response):
-        json_data = json.loads(response.text)
+        data = json.loads(response.text)
         yield self.build_file_from_response(
             response,
-            data=json.dumps(json_data['data']).encode(),
+            data=json.dumps(data['data']).encode(),
             data_type='record_package_list'
         )
diff --git a/kingfisher_scrapy/spiders/moldova.py b/kingfisher_scrapy/spiders/moldova.py
index fb146a61..2883379c 100644
--- a/kingfisher_scrapy/spiders/moldova.py
+++ b/kingfisher_scrapy/spiders/moldova.py
@@ -1,62 +1,40 @@
 import json
 
-import scrapy
+from kingfisher_scrapy.base_spider import SimpleSpider
+from kingfisher_scrapy.util import components, handle_error, join, parameters, replace_parameter
 
-from kingfisher_scrapy.base_spider import BaseSpider
-from kingfisher_scrapy.util import handle_error
 
-
-class Moldova(BaseSpider):
+class Moldova(SimpleSpider):
     name = 'moldova'
-
-    endpoints = {"budgets": "https://public.mtender.gov.md/budgets/",
-                 # From https://github.com/open-contracting/kingfisher-collect/issues/192#issuecomment-529928683
-                 # The /tenders/plans endpoint appeared to return exactly the same data as the /tenders endpoint except
-                 # that when given an OCID parameter it returned an error message. It may be that /tenders/plans just
-                 # lists a subset of /tenders but this isn't clear.
-                 # "plans": "https://public.mtender.gov.md/tenders/plan/",
-                 "tenders": "https://public.mtender.gov.md/tenders/"}
+    data_type = 'record_package'
 
     def start_requests(self):
-        for endpoint, url in self.endpoints.items():
-            yield scrapy.Request(
-                url,
-                meta={'kf_filename': 'meta-{}-start.json'.format(endpoint), 'endpoint': endpoint, 'data': False}
-            )
+        endpoints = {
+            'budgets': 'https://public.mtender.gov.md/budgets/',
+            # From https://github.com/open-contracting/kingfisher-collect/issues/192#issuecomment-529928683
+            # The /tenders/plans endpoint appeared to return exactly the same data as the /tenders endpoint except
+            # that when given an OCID parameter it returned an error message. It may be that /tenders/plans just
+            # lists a subset of /tenders but this isn't clear.
+            # 'plans': 'https://public.mtender.gov.md/tenders/plan/',
+            'tenders': 'https://public.mtender.gov.md/tenders/',
+        }
+
+        for endpoint, url in endpoints.items():
+            yield self.build_request(url, formatter=components(-1), callback=self.parse_list)
 
     @handle_error
-    def parse(self, response):
-        if response.request.meta['data']:
-            yield self.build_file_from_response(response, data_type='record_package')
-        else:
-            self.build_file_from_response(response)
-            json_data = json.loads(response.text)
-            offset = json_data.get('offset')
-            # not having an offset in the data means the data has come to an end.
-            if not offset:
-                return
-
-            endpoint = response.request.meta['endpoint']
-            endpoint_url = self.endpoints[endpoint]
-
-            for data in json_data.get('data', []):
-                yield scrapy.Request(
-                    endpoint_url + data['ocid'],
-                    meta={
-                        'kf_filename': 'data-{}-{}.json'.format(endpoint, data['ocid']),
-                        'endpoint': endpoint,
-                        'data': True,
-                    }
-                )
-
-            if self.sample:
-                return
-
-            yield scrapy.Request(
-                endpoint_url + '?offset=' + offset,
-                meta={
-                    'kf_filename': 'meta-{}-{}.json'.format(endpoint, offset),
-                    'endpoint': endpoint,
-                    'data': False,
-                }
-            )
+    def parse_list(self, response):
+        data = json.loads(response.text)
+        # The last page returns an empty JSON object.
+        if not data:
+            return
+
+        for item in data['data']:
+            url = replace_parameter(response.request.url, 'offset', None) + item['ocid']
+            yield self.build_request(url, formatter=components(-2))
+
+        if self.sample:
+            return
+
+        url = replace_parameter(response.request.url, 'offset', data['offset'])
+        yield self.build_request(url, formatter=join(components(-1), parameters('offset')), callback=self.parse_list)
diff --git a/kingfisher_scrapy/spiders/moldova_old.py b/kingfisher_scrapy/spiders/moldova_old.py
index a8324c21..db50f1d7 100644
--- a/kingfisher_scrapy/spiders/moldova_old.py
+++ b/kingfisher_scrapy/spiders/moldova_old.py
@@ -1,6 +1,5 @@
-import scrapy
-
 from kingfisher_scrapy.base_spider import SimpleSpider
+from kingfisher_scrapy.util import components, date_range_by_year
 
 
 class MoldovaOld(SimpleSpider):
@@ -8,14 +7,12 @@ class MoldovaOld(SimpleSpider):
     data_type = 'release_package'
 
     def start_requests(self):
+        pattern = 'http://opencontracting.date.gov.md/ocds-api/year/{}'
+
+        start = 2012
+        stop = 2018
         if self.sample:
-            yield scrapy.Request(
-                'http://opencontracting.date.gov.md/ocds-api/year/2017',
-                meta={'kf_filename': 'sample.json'}
-            )
-        else:
-            for year in range(2012, 2018):
-                yield scrapy.Request(
-                    'http://opencontracting.date.gov.md/ocds-api/year/%d' % year,
-                    meta={'kf_filename': 'year-%d.json' % year}
-                )
+            start = 2018
+
+        for year in date_range_by_year(start, stop):
+            yield self.build_request(pattern.format(year), formatter=components(-1))
diff --git a/kingfisher_scrapy/spiders/nepal_dhangadhi.py b/kingfisher_scrapy/spiders/nepal_dhangadhi.py
index 1b0f635a..2e0fbdf6 100644
--- a/kingfisher_scrapy/spiders/nepal_dhangadhi.py
+++ b/kingfisher_scrapy/spiders/nepal_dhangadhi.py
@@ -1,10 +1,9 @@
-import hashlib
 import json
 
 import scrapy
 
 from kingfisher_scrapy.base_spider import SimpleSpider
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import components, handle_error
 
 
 class NepalDhangadhi(SimpleSpider):
@@ -20,14 +19,9 @@ def start_requests(self):
 
     @handle_error
     def parse_list(self, response):
-        url = 'https://admin.ims.susasan.org/ocds/json/dhangadhi-{}.json'
-        json_data = json.loads(response.text)
-        fiscal_years = json_data['data']['fiscal_years']
-        for item in fiscal_years:
-            fy = item['name']
-            yield scrapy.Request(
-                url.format(fy),
-                meta={'kf_filename': hashlib.md5((url + fy).encode('utf-8')).hexdigest() + '.json'},
-            )
+        pattern = 'https://admin.ims.susasan.org/ocds/json/dhangadhi-{}.json'
+        data = json.loads(response.text)
+        for item in data['data']['fiscal_years']:
+            yield self.build_request(pattern.format(item['name']), formatter=components(-1))
             if self.sample:
                 break
diff --git a/kingfisher_scrapy/spiders/nepal_portal.py b/kingfisher_scrapy/spiders/nepal_portal.py
index 3cc70953..3b91a876 100644
--- a/kingfisher_scrapy/spiders/nepal_portal.py
+++ b/kingfisher_scrapy/spiders/nepal_portal.py
@@ -1,9 +1,7 @@
-import datetime
-import hashlib
-
-import scrapy
+from datetime import date
 
 from kingfisher_scrapy.base_spider import SimpleSpider
+from kingfisher_scrapy.util import components, date_range_by_year
 
 
 class NepalPortal(SimpleSpider):
@@ -11,18 +9,14 @@ class NepalPortal(SimpleSpider):
     data_type = 'release_package'
 
     def start_requests(self):
+        pattern = 'http://ppip.gov.np/bulk-download/{}'
+
         if self.sample:
-            current_year = 2018
-            end_year = 2018
+            start = 2018
+            stop = 2018
         else:
-            current_year = 2013
-            now = datetime.datetime.now()
-            end_year = now.year
+            start = 2012
+            stop = date.today().year  # HTTP 500 after 2018
 
-        while current_year <= end_year:
-            url = 'http://ppip.gov.np/bulk-download/{}'.format(current_year)
-            yield scrapy.Request(
-                url,
-                meta={'kf_filename': hashlib.md5(url.encode('utf-8')).hexdigest() + '.json'}
-            )
-            current_year += 1
+        for year in date_range_by_year(start, stop):
+            yield self.build_request(pattern.format(year), formatter=components(-1))
diff --git a/kingfisher_scrapy/spiders/nigeria_portal.py b/kingfisher_scrapy/spiders/nigeria_portal.py
index 82eac852..fc8509f0 100644
--- a/kingfisher_scrapy/spiders/nigeria_portal.py
+++ b/kingfisher_scrapy/spiders/nigeria_portal.py
@@ -1,5 +1,3 @@
-import hashlib
-
 import scrapy
 
 from kingfisher_scrapy.base_spider import SimpleSpider
@@ -16,7 +14,7 @@ class NigeriaPortal(SimpleSpider):
     def start_requests(self):
         yield scrapy.Request(
             'http://nocopo.bpp.gov.ng/OpenData.aspx',
-            meta={'kf_filename': 'list.html'},
+            meta={'kf_filename': 'form.html'},
             callback=self.parse_list
         )
 
@@ -36,8 +34,4 @@ def parse_list(self, response):
                 if self.sample:
                     break
 
-        yield scrapy.FormRequest.from_response(
-            response,
-            formdata=formdata,
-            meta={'kf_filename': hashlib.md5(response.url.encode('utf-8')).hexdigest() + '.json'}
-        )
+        yield scrapy.FormRequest.from_response(response, formdata=formdata, meta={'kf_filename': 'all.json'})
diff --git a/kingfisher_scrapy/spiders/openopps.py b/kingfisher_scrapy/spiders/openopps.py
index c3399af9..e5d6ee32 100644
--- a/kingfisher_scrapy/spiders/openopps.py
+++ b/kingfisher_scrapy/spiders/openopps.py
@@ -1,4 +1,3 @@
-import hashlib
 import json
 from datetime import datetime, timedelta
 from math import ceil
@@ -7,6 +6,7 @@
 
 from kingfisher_scrapy.base_spider import BaseSpider
 from kingfisher_scrapy.exceptions import AuthenticationError
+from kingfisher_scrapy.util import parameters
 
 
 class OpenOpps(BaseSpider):
@@ -40,9 +40,8 @@ class OpenOpps(BaseSpider):
     reauthenticating = False  # flag for request a new token
     start_time = None
 
-    base_page_url = \
-        'https://api.openopps.com/api/ocds/?' \
-        'format={}&ordering={}&page_size={}&releasedate__gte={{}}&releasedate__lte={{}}'
+    base_page_url = 'https://api.openopps.com/api/ocds/?format=json&ordering=releasedate&page_size=1000&' \
+                    'releasedate__gte={}&releasedate__lte={}'
 
     custom_settings = {
         'DOWNLOADER_MIDDLEWARES': {
@@ -82,7 +81,7 @@ def parse_access_token(self, response):
             r = json.loads(response.text)
             token = r.get('token')
             if token:
-                self.logger.info('New access token: {}'.format(token))
+                self.logger.info(f'New access token: {token}')
                 self.access_token = 'JWT ' + token
                 self.start_time = datetime.now()
                 # If the request is initial authentication, start requests
@@ -92,19 +91,15 @@ def parse_access_token(self, response):
                 self.reauthenticating = False
             else:
                 self.logger.error(
-                    'Authentication failed. Status code: {}. {}'.format(response.status, response.text))
+                    f'Authentication failed. Status code: {response.status}. {response.text}')
                 raise AuthenticationError()
         else:
             self.logger.error(
-                'Authentication failed. Status code: {}. {}'.format(response.status, response.text))
+                f'Authentication failed. Status code: {response.status}. {response.text}')
             raise AuthenticationError()
 
     def start_requests_pages(self):
-        page_size = 1000
-        page_format = 'json'
-        ordering = 'releasedate'
         search_h = 24  # start splitting one day search
-        self.base_page_url = self.base_page_url.format(page_format, ordering, page_size)
 
         # Case if we want to download a sample
         if self.sample:
@@ -130,11 +125,10 @@ def start_requests_pages(self):
                     yield from self.request_range_per_day(start_date, end_date, search_h)
 
     def request_range(self, start_date, end_date, search_h):
-        url = self.base_page_url.format(start_date, end_date)
-        return scrapy.Request(
-            url,
+        return self.build_request(
+            self.base_page_url.format(start_date, end_date),
+            formatter=parameters('releasedate__gte', 'releasedate__lte'),
             meta={
-                'kf_filename': hashlib.md5(url.encode('utf-8')).hexdigest() + '.json',
                 'release_date': start_date,
                 'search_h': search_h,
             },
@@ -171,10 +165,10 @@ def parse(self, response):
 
                 next_url = results.get('next')
                 if next_url:
-                    yield scrapy.Request(
+                    yield self.build_request(
                         next_url,
+                        formatter=parameters('releasedate__gte', 'releasedate__lte', 'page'),
                         meta={
-                            'kf_filename': hashlib.md5(next_url.encode('utf-8')).hexdigest() + '.json',
                             'release_date': release_date,
                             'search_h': search_h,
                         },
@@ -184,7 +178,7 @@ def parse(self, response):
                 # Tells if we have to re-authenticate before the token expires
                 time_diff = datetime.now() - self.start_time
                 if not self.reauthenticating and time_diff.total_seconds() > self.request_time_limit * 60:
-                    self.logger.info('Time_diff: {}'.format(time_diff.total_seconds()))
+                    self.logger.info(f'Time_diff: {time_diff.total_seconds()}')
                     self.reauthenticating = True
                     yield scrapy.Request(
                         'https://api.openopps.com/api/api-token-auth/',
@@ -220,13 +214,12 @@ def parse(self, response):
                 if len(start_hour_list) != len(end_hour_list):
                     end_hour_list.append(last_hour)
 
-                self.logger.info('Changing filters, split in {}: {}.'.format(parts, response.request.url))
+                self.logger.info(f'Changing filters, split in {parts}: {response.request.url}.')
                 for i in range(len(start_hour_list)):
-                    url = self.base_page_url.format(start_hour_list[i], end_hour_list[i])
-                    yield scrapy.Request(
-                        url,
+                    yield self.build_request(
+                        self.base_page_url.format(start_hour_list[i], end_hour_list[i]),
+                        formatter=parameters('releasedate__gte', 'releasedate__lte'),
                         meta={
-                            'kf_filename': hashlib.md5(url.encode('utf-8')).hexdigest() + '.json',
                             'release_date': start_hour_list[i],  # release_date with star hour
                             'last_hour': end_hour_list[i],  # release_date with last hour
                             'search_h': split_h,  # new search range
@@ -237,7 +230,7 @@ def parse(self, response):
             # Message for pages that exceed the 10,000 search results in the range of one hour
             # These are pages with status 500 and 'page=11' in the URL request
             if response.status == 500 and response.request.url.count("page=11"):
-                self.logger.info('Status: {}. Results exceeded in a range of one hour, we save the '
-                                 'first 10,000 data for: {}'.format(response.status, response.request.url))
+                self.logger.info(f'Status: {response.status}. Results exceeded in a range of one hour, we save the '
+                                 f'first 10,000 data for: {response.request.url}')
             else:
                 yield self.build_file_error_from_response(response)
diff --git a/kingfisher_scrapy/spiders/paraguay_dncp_base.py b/kingfisher_scrapy/spiders/paraguay_dncp_base.py
index 4c6e265f..30580bfb 100644
--- a/kingfisher_scrapy/spiders/paraguay_dncp_base.py
+++ b/kingfisher_scrapy/spiders/paraguay_dncp_base.py
@@ -6,7 +6,7 @@
 
 from kingfisher_scrapy.base_spider import SimpleSpider
 from kingfisher_scrapy.exceptions import AuthenticationError
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import components, handle_error, parameters
 
 
 class ParaguayDNCPBaseSpider(SimpleSpider):
@@ -22,8 +22,8 @@ class ParaguayDNCPBaseSpider(SimpleSpider):
     last_request = None
     request_time_limit = 13  # in minutes
     base_url = 'https://contrataciones.gov.py/datos/api/v3/doc'
-    base_page_url = '{}/search/processes?fecha_desde=2010-01-01'.format(base_url)
-    auth_url = '{}/oauth/token'.format(base_url)
+    base_page_url = f'{base_url}/search/processes?fecha_desde=2010-01-01'
+    auth_url = f'{base_url}/oauth/token'
     request_token = None
     max_attempts = 10
     data_type = None
@@ -54,10 +54,10 @@ def start_requests(self):
             from_date = self.from_date.strftime(self.date_format)
             self.base_page_url = '{}/search/processes?tipo_fecha=fecha_release&fecha_desde={}'\
                 .format(self.base_url, from_date)
-        yield scrapy.Request(
+        yield self.build_request(
             self.base_page_url,
+            formatter=parameters('fecha_desde'),
             meta={
-                'kf_filename': '{}-1.json'.format(from_date),
                 'from_date': from_date,
             },
             # send duplicate requests when the token expired and in the continuation of last_request saved.
@@ -69,7 +69,7 @@ def request_access_token(self):
         """ Requests a new access token """
         attempt = 0
         self.start_time = datetime.now()
-        self.logger.info('Requesting access token, attempt {} of {}'.format(attempt + 1, self.max_attempts))
+        self.logger.info(f'Requesting access token, attempt {attempt + 1} of {self.max_attempts}')
 
         return scrapy.Request(
             self.auth_url,
@@ -87,7 +87,7 @@ def parse_access_token(self, response):
             r = json.loads(response.text)
             token = r.get('access_token')
             if token:
-                self.logger.info('New access token: {}'.format(token))
+                self.logger.info(f'New access token: {token}')
                 self.access_token = token
                 # continue scraping where it stopped after getting the token
                 yield self.last_request
@@ -98,10 +98,7 @@ def parse_access_token(self, response):
                     self.auth_failed = True
                     raise AuthenticationError()
                 else:
-                    self.logger.info('Requesting access token, attempt {} of {}'.format(
-                        attempt + 1,
-                        self.max_attempts)
-                    )
+                    self.logger.info(f'Requesting access token, attempt {attempt + 1} of {self.max_attempts}')
                     return scrapy.Request(
                         self.auth_url,
                         method='POST',
@@ -113,7 +110,7 @@ def parse_access_token(self, response):
                         priority=1000
                     )
         else:
-            self.logger.error('Authentication failed. Status code: {}'.format(response.status))
+            self.logger.error(f'Authentication failed. Status code: {response.status}')
             self.auth_failed = True
             raise AuthenticationError()
 
@@ -121,18 +118,14 @@ def parse_access_token(self, response):
     def parse_pages(self, response):
         content = json.loads(response.text)
         for url in self.get_files_to_download(content):
-            yield scrapy.Request(
-                url,
-                dont_filter=True,
-                meta={'kf_filename': url.split('/')[-1] + '.json'}
-            )
+            yield self.build_request(url, formatter=components(-1), dont_filter=True)
         pagination = content['pagination']
         if pagination['current_page'] < pagination['total_pages'] and not self.sample:
             page = pagination['current_page'] + 1
-            url = '{}&page={}'.format(self.base_page_url, page)
-            yield scrapy.Request(
+            url = f'{self.base_page_url}&page={page}'
+            yield self.build_request(
                 url,
-                meta={'kf_filename': '{}-{}.json'.format(response.request.meta['from_date'], page)},
+                formatter=parameters('fecha_desde', 'page'),
                 dont_filter=True,
                 callback=self.parse_pages
             )
@@ -148,5 +141,5 @@ def expires_soon(self, time_diff):
         """
         if time_diff.total_seconds() < ParaguayDNCPBaseSpider.request_time_limit * 60:
             return False
-        self.logger.info('Time_diff: {}'.format(time_diff.total_seconds()))
+        self.logger.info(f'Time_diff: {time_diff.total_seconds()}')
         return True
diff --git a/kingfisher_scrapy/spiders/paraguay_dncp_records.py b/kingfisher_scrapy/spiders/paraguay_dncp_records.py
index 6325cef6..721e666a 100644
--- a/kingfisher_scrapy/spiders/paraguay_dncp_records.py
+++ b/kingfisher_scrapy/spiders/paraguay_dncp_records.py
@@ -20,4 +20,4 @@ class ParaguayDNCPRecords(ParaguayDNCPBaseSpider):
 
     def get_files_to_download(self, content):
         for record in content['records']:
-            yield '{}/ocds/record/{}'.format(self.base_url, record['ocid'])
+            yield f"{self.base_url}/ocds/record/{record['ocid']}"
diff --git a/kingfisher_scrapy/spiders/paraguay_hacienda.py b/kingfisher_scrapy/spiders/paraguay_hacienda.py
index d10ebb98..522dbde9 100644
--- a/kingfisher_scrapy/spiders/paraguay_hacienda.py
+++ b/kingfisher_scrapy/spiders/paraguay_hacienda.py
@@ -5,7 +5,7 @@
 
 from kingfisher_scrapy.base_spider import BaseSpider
 from kingfisher_scrapy.exceptions import AuthenticationError
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import components, handle_error, parameters
 
 
 class ParaguayHacienda(BaseSpider):
@@ -42,10 +42,10 @@ def from_crawler(cls, crawler, *args, **kwargs):
     def start_requests(self):
         # Paraguay Hacienda has a service that return all the ids that we need to get the releases packages
         # so we first iterate over this list that is paginated
-        yield scrapy.Request(
+        yield self.build_request(
             self.base_list_url.format(1),
+            formatter=parameters('page'),
             meta={
-                'kf_filename': 'list-1.json',
                 'meta': True,
                 'first': True,
             },
@@ -56,16 +56,16 @@ def start_requests(self):
     @handle_error
     def parse(self, response):
         data = json.loads(response.text)
-        base_url = 'https://datos.hacienda.gov.py:443/odmh-api-v1/rest/api/v1/ocds/release-package/{}'
+        pattern = 'https://datos.hacienda.gov.py:443/odmh-api-v1/rest/api/v1/ocds/release-package/{}'
 
         # If is the first URL, we need to iterate over all the pages to get all the process ids to query
         if response.request.meta['first'] and not self.sample:
-            total_pages = data['meta']['totalPages']
-            for page in range(2,  total_pages+1):
-                yield scrapy.Request(
+            total = data['meta']['totalPages']
+            for page in range(2,  total + 1):
+                yield self.build_request(
                     self.base_list_url.format(page),
+                    formatter=parameters('page'),
                     meta={
-                        'kf_filename': 'list-{}.json'.format(page),
                         'meta': True,
                         'first': False,
                     },
@@ -82,10 +82,10 @@ def parse(self, response):
             for row in data['results']:
                 if row['idLlamado'] and row['idLlamado'] not in self.release_ids:
                     self.release_ids.append(row['idLlamado'])
-                    yield scrapy.Request(
-                        base_url.format(row['idLlamado']),
+                    yield self.build_request(
+                        pattern.format(row['idLlamado']),
+                        formatter=components(-1),
                         meta={
-                            'kf_filename': 'release-{}.json'.format(row['idLlamado']),
                             'meta': False,
                             'first': False,
                         },
@@ -98,7 +98,7 @@ def request_access_token(self):
         """ Requests a new access token """
         attempt = 0
         self.start_time = datetime.now()
-        self.logger.info('Requesting access token, attempt {} of {}'.format(attempt + 1, self.max_attempts))
+        self.logger.info(f'Requesting access token, attempt {attempt + 1} of {self.max_attempts}')
         payload = {"clientSecret": self.client_secret}
 
         return scrapy.Request(
@@ -117,7 +117,7 @@ def parse_access_token(self, response):
             r = json.loads(response.text)
             token = r.get('accessToken')
             if token:
-                self.logger.info('New access token: {}'.format(token))
+                self.logger.info(f'New access token: {token}')
                 self.access_token = 'Bearer ' + token
                 # continue scraping where it stopped after getting the token
                 yield self.last_request
@@ -128,10 +128,7 @@ def parse_access_token(self, response):
                     self.auth_failed = True
                     raise AuthenticationError()
                 else:
-                    self.logger.info('Requesting access token, attempt {} of {}'.format(
-                        attempt + 1,
-                        self.max_attempts)
-                    )
+                    self.logger.info(f'Requesting access token, attempt {attempt + 1} of {self.max_attempts}')
                     return scrapy.Request(
                         "https://datos.hacienda.gov.py:443/odmh-api-v1/rest/api/v1/auth/token",
                         method='POST',
@@ -143,7 +140,7 @@ def parse_access_token(self, response):
                         priority=1000
                     )
         else:
-            self.logger.error('Authentication failed. Status code: {}'.format(response.status))
+            self.logger.error(f'Authentication failed. Status code: {response.status}')
             self.auth_failed = True
             raise AuthenticationError()
 
@@ -153,5 +150,5 @@ def expires_soon(self, time_diff):
         """
         if time_diff.total_seconds() < ParaguayHacienda.request_time_limit * 60:
             return False
-        self.logger.info('Time_diff: {}'.format(time_diff.total_seconds()))
+        self.logger.info(f'Time_diff: {time_diff.total_seconds()}')
         return True
diff --git a/kingfisher_scrapy/spiders/portugal.py b/kingfisher_scrapy/spiders/portugal.py
index 7a3b5f5c..fdf25ba7 100644
--- a/kingfisher_scrapy/spiders/portugal.py
+++ b/kingfisher_scrapy/spiders/portugal.py
@@ -1,10 +1,9 @@
-import hashlib
 import json
 
 import scrapy
 
 from kingfisher_scrapy.base_spider import ZipSpider
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import components, handle_error
 
 
 class Portugal(ZipSpider):
@@ -17,26 +16,16 @@ class Portugal(ZipSpider):
     download_timeout = 9999
 
     def start_requests(self):
-        url = 'https://dados.gov.pt/api/1/datasets/?q=ocds&organization={}&page_size={}'
-        id = '5ae97fa2c8d8c915d5faa3bf'
-        page_size = 20
-        yield scrapy.Request(
-            url.format(id, page_size),
-            meta={'kf_filename': 'list.json'},
-            callback=self.parse_list
-        )
+        # A CKAN API JSON response.
+        url = 'https://dados.gov.pt/api/1/datasets/?q=ocds&organization=5ae97fa2c8d8c915d5faa3bf&page_size=20'
+        yield scrapy.Request(url, meta={'kf_filename': 'page-1.json'}, callback=self.parse_list)
 
     @handle_error
     def parse_list(self, response):
-        datas = json.loads(response.text)
-        for data in datas['data']:
-            for resource in data['resources']:
+        data = json.loads(response.text)
+        for item in data['data']:
+            for resource in item['resources']:
                 description = resource['description']
-                url = resource['url']
-                if description.count("OCDS") or description.count("ocds"):
-                    yield scrapy.Request(
-                        url,
-                        meta={'kf_filename': hashlib.md5(url.encode('utf-8')).hexdigest() + '.json'}
-                    )
-                    if self.sample:
-                        break
+                if description and 'ocds' in description.lower():
+                    # Presently, only one URL matches.
+                    yield self.build_request(resource['url'], formatter=components(-2))
diff --git a/kingfisher_scrapy/spiders/scotland.py b/kingfisher_scrapy/spiders/scotland.py
index 76b365dc..e8cb5973 100644
--- a/kingfisher_scrapy/spiders/scotland.py
+++ b/kingfisher_scrapy/spiders/scotland.py
@@ -1,53 +1,51 @@
-import datetime
-
-import scrapy
+from datetime import date, timedelta
 
 from kingfisher_scrapy.base_spider import SimpleSpider
+from kingfisher_scrapy.util import parameters
 
 
 class Scotland(SimpleSpider):
     name = 'scotland'
     data_type = 'release_package'
 
-    notice_types = [
-        1,  # OJEU - F1 - Prior Information Notice
-        2,  # OJEU - F2 - Contract Notice
-        3,  # OJEU - F3 - Contract Award Notice
-        4,  # OJEU - F4 - Prior Information Notice(Utilities)
-        5,  # OJEU - F5 - Contract Notice(Utilities)
-        6,  # OJEU - F6 - Contract Award Notice(Utilities)
-        7,  # OJEU - F7 - Qualification Systems(Utilities)
-        12,  # OJEU - F12 - Design Contest Notice
-        13,  # OJEU - F13 - Results Of Design Contest
-        14,  # OJEU - F14 - Corrigendum
-        15,  # OJEU - F15 - Voluntary Ex Ante Transparency Notice
-        20,  # OJEU - F20 - Modification Notice
-        21,  # OJEU - F21 - Social And other Specific Services(Public Contracts)
-        22,  # OJEU - F22 - Social And other Specific Services(Utilities)
-        23,  # OJEU - F23 - Social And other Specific Services(Concessions)
-        24,  # OJEU - F24 - Concession Notice
-        25,  # OJEU - F25 - Concession Award Notice
-        101,  # Site Notice - Website Contract Notice
-        102,  # Site Notice - Website Prior Information Notice
-        103,  # Site Notice - Website Contract Award Notice
-        104,  # Site Notice - Quick Quote Award
-    ]
-
     def start_requests(self):
-        format_string = 'https://api.publiccontractsscotland.gov.uk/v1/Notices?dateFrom={}&outputType=1&noticeType={}'
+        pattern = 'https://api.publiccontractsscotland.gov.uk/v1/Notices?dateFrom={}&outputType=1&noticeType={}'
+
+        notice_types = [
+            1,  # OJEU - F1 - Prior Information Notice
+            2,  # OJEU - F2 - Contract Notice
+            3,  # OJEU - F3 - Contract Award Notice
+            4,  # OJEU - F4 - Prior Information Notice(Utilities)
+            5,  # OJEU - F5 - Contract Notice(Utilities)
+            6,  # OJEU - F6 - Contract Award Notice(Utilities)
+            7,  # OJEU - F7 - Qualification Systems(Utilities)
+            12,  # OJEU - F12 - Design Contest Notice
+            13,  # OJEU - F13 - Results Of Design Contest
+            14,  # OJEU - F14 - Corrigendum
+            15,  # OJEU - F15 - Voluntary Ex Ante Transparency Notice
+            20,  # OJEU - F20 - Modification Notice
+            21,  # OJEU - F21 - Social And other Specific Services(Public Contracts)
+            22,  # OJEU - F22 - Social And other Specific Services(Utilities)
+            23,  # OJEU - F23 - Social And other Specific Services(Concessions)
+            24,  # OJEU - F24 - Concession Notice
+            25,  # OJEU - F25 - Concession Award Notice
+            101,  # Site Notice - Website Contract Notice
+            102,  # Site Notice - Website Prior Information Notice
+            103,  # Site Notice - Website Contract Award Notice
+            104,  # Site Notice - Quick Quote Award
+        ]
+
+        now = date.today()
 
-        now = datetime.datetime.today()
-        if self.sample:
-            marker = now - datetime.timedelta(days=14)
-            for notice_type in self.notice_types:
-                yield scrapy.Request(format_string.format(marker, notice_type),
-                                     meta={'kf_filename': 'sample_{}.json'.format(notice_type)})
-        else:
-            # It's meant to go back a year, but in testing it seemed to be year minus one day!
-            marker = now - datetime.timedelta(days=364)
-            while marker <= now:
-                datestring = '{:04d}-{:02d}-{:02d}'.format(marker.year, marker.month, marker.day)
-                for notice_type in self.notice_types:
-                    yield scrapy.Request(format_string.format(datestring, notice_type),
-                                         meta={'kf_filename': '{}_type_{}.json'.format(datestring, notice_type)})
-                marker = marker + datetime.timedelta(days=14)
+        # It's meant to go back a year, but in testing it seemed to be year minus one day!
+        marker = now - timedelta(days=364)
+        while marker <= now:
+            datestring = '{:04d}-{:02d}-{:02d}'.format(marker.year, marker.month, marker.day)
+            for notice_type in notice_types:
+                yield self.build_request(
+                    pattern.format(datestring, notice_type),
+                    formatter=parameters('noticeType', 'dateFrom')
+                )
+            marker = marker + timedelta(days=14)
+            if self.sample:
+                break
diff --git a/kingfisher_scrapy/spiders/uganda_releases.py b/kingfisher_scrapy/spiders/uganda_releases.py
index 1e8f5305..3a629c59 100644
--- a/kingfisher_scrapy/spiders/uganda_releases.py
+++ b/kingfisher_scrapy/spiders/uganda_releases.py
@@ -1,10 +1,9 @@
-import hashlib
 import json
 
 import scrapy
 
 from kingfisher_scrapy.base_spider import SimpleSpider
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import components, handle_error, join, parameters
 
 
 class Uganda(SimpleSpider):
@@ -16,54 +15,36 @@ class Uganda(SimpleSpider):
     def start_requests(self):
         yield scrapy.Request(
             'https://gpp.ppda.go.ug/adminapi/public/api/pdes',
-            meta={'kf_filename': 'start_requests'},
-            callback=self.parse_pages
+            meta={'kf_filename': 'page-1.json'},
+            callback=self.parse_list
         )
 
     @handle_error
-    def parse_pages(self, response):
-        url_pdes = 'https://gpp.ppda.go.ug/adminapi/public/api/pdes?page={}'
+    def parse_list(self, response):
+        pattern = 'https://gpp.ppda.go.ug/adminapi/public/api/pdes?page={}'
 
         if self.sample:
-            total_pages = 1
+            total = 1
         else:
-            json_data = json.loads(response.text)
-            total_pages = json_data['data']['last_page']
+            data = json.loads(response.text)
+            total = data['data']['last_page']
 
-        for page_number in range(total_pages):
-            yield scrapy.Request(
-                url_pdes.format(page_number + 1),
-                meta={'kf_filename': 'pages_requests'},
-                callback=self.parse_data
-            )
+        for page in range(2, total + 1):
+            yield self.build_request(pattern.format(page), formatter=parameters('page'), callback=self.parse_data)
 
     @handle_error
     def parse_data(self, response):
-        url = 'https://gpp.ppda.go.ug/adminapi/public/api/open-data/v1/releases/{}?fy={}&pde={}'
-        tags = ['planning', 'tender', 'award', 'contract']
-        pdes_fdy_checks = []
-
-        json_data = json.loads(response.text)
-        list_pdes = json_data['data']['data']
-
-        for pdes in list_pdes:
-            pde_plans = pdes['procurement_plans']
-
-            for plans in pde_plans:
-                financial_year = plans['financial_year']
-                procurement_entity_id = plans['pde_id']
-                pdes_fdy = financial_year + '&' + procurement_entity_id
-
-                if pdes_fdy not in pdes_fdy_checks:
-                    pdes_fdy_checks.append(pdes_fdy)
-
-                    for tag in tags:
-                        yield scrapy.Request(
-                            url.format(tag, financial_year, procurement_entity_id),
-                            meta={'kf_filename': hashlib.md5(
-                                (url + str(pdes_fdy + tag)).encode('utf-8')).hexdigest() + '.json'}
-                        )
-                        if self.sample:
-                            break
+        pattern = 'https://gpp.ppda.go.ug/adminapi/public/api/open-data/v1/releases/{}?fy={}&pde={}'
+
+        data = json.loads(response.text)
+        for pdes in data['data']['data']:
+            for plans in pdes['procurement_plans']:
+                for tag in ('planning', 'tender', 'award', 'contract'):
+                    yield self.build_request(
+                        pattern.format(tag, plans['financial_year'], plans['pde_id']),
+                        formatter=join(components(-1), parameters('fy', 'pde'))
+                    )
                     if self.sample:
                         break
+                if self.sample:
+                    break
diff --git a/kingfisher_scrapy/spiders/uk_contracts_finder.py b/kingfisher_scrapy/spiders/uk_contracts_finder.py
index 333cc89c..baf3dce3 100644
--- a/kingfisher_scrapy/spiders/uk_contracts_finder.py
+++ b/kingfisher_scrapy/spiders/uk_contracts_finder.py
@@ -1,28 +1,25 @@
 import json
 
-import scrapy
-
 from kingfisher_scrapy.base_spider import BaseSpider
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import components, handle_error, parameters, replace_parameter
 
 
 class UKContractsFinder(BaseSpider):
     name = 'uk_contracts_finder'
-    base_url = 'https://www.contractsfinder.service.gov.uk/Published/Notices/OCDS/Search?order=asc&page=%d'
+    data_type = 'release_package_list_in_results'
+    encoding = 'iso-8859-1'
 
     def start_requests(self):
-        yield scrapy.Request(self.base_url % 1, meta={'kf_filename': 'page1.json'})
+        url = 'https://www.contractsfinder.service.gov.uk/Published/Notices/OCDS/Search?order=asc&page=1'
+        yield self.build_request(url, formatter=parameters('page'), callback=self.parse_list)
 
     @handle_error
-    def parse(self, response):
-        yield self.build_file_from_response(
-            response,
-            data_type='release_package_list_in_results',
-            encoding='iso-8859-1'
-        )
+    def parse_list(self, response):
+        yield from self.parse(response)
 
-        if not self.sample and response.request.meta['kf_filename'] == 'page1.json':
-            json_data = json.loads(response.text)
-            last_page = json_data['maxPage']
-            for page in range(1, last_page + 1):
-                yield scrapy.Request(self.base_url % page, meta={'kf_filename': 'page%d.json' % page})
+        if not self.sample:
+            data = json.loads(response.text)
+            total = data['maxPage']
+            for page in range(2, total + 1):
+                url = replace_parameter(response.request.url, 'page', page)
+                yield self.build_request(url, formatter=components('page'))
diff --git a/kingfisher_scrapy/spiders/uk_fts.py b/kingfisher_scrapy/spiders/uk_fts.py
index a876784e..6c172dbf 100644
--- a/kingfisher_scrapy/spiders/uk_fts.py
+++ b/kingfisher_scrapy/spiders/uk_fts.py
@@ -1,16 +1,15 @@
 import scrapy
 
 from kingfisher_scrapy.base_spider import LinksSpider
+from kingfisher_scrapy.util import parameters
 
 
 class UKContractsFinder(LinksSpider):
     name = 'uk_fts'
     data_type = 'release_package_in_ocdsReleasePackage_in_list_in_results'
+    next_page_formatter = parameters('cursor')
 
     def start_requests(self):
-        yield scrapy.Request(
-            # This URL was provided by the publisher and is not the production URL.
-            url='https://enoticetest.service.xgov.uk/api/1.0/ocdsReleasePackages',
-            meta={'kf_filename': 'start.json'},
-            headers={'Accept': 'application/json'},
-        )
+        # This URL was provided by the publisher and is not the production URL.
+        url = 'https://enoticetest.service.xgov.uk/api/1.0/ocdsReleasePackages'
+        yield scrapy.Request(url, meta={'kf_filename': 'start.json'}, headers={'Accept': 'application/json'})
diff --git a/kingfisher_scrapy/spiders/uruguay_base.py b/kingfisher_scrapy/spiders/uruguay_base.py
index f1d6152a..d554fe89 100644
--- a/kingfisher_scrapy/spiders/uruguay_base.py
+++ b/kingfisher_scrapy/spiders/uruguay_base.py
@@ -1,30 +1,19 @@
-import hashlib
-from datetime import date, timedelta
-
-import scrapy
+from datetime import date
 
 from kingfisher_scrapy.base_spider import SimpleSpider
+from kingfisher_scrapy.util import components, date_range_by_month
 
 
 class UruguayBase(SimpleSpider):
     download_delay = 0.9
 
     def start_requests(self):
-        base_url = 'http://comprasestatales.gub.uy/ocds/rss/{year:d}/{month:02d}'
+        url = 'http://comprasestatales.gub.uy/ocds/rss/{0.year:d}/{0.month:02d}'
 
-        current_date = date(2017, 11, 1)
+        start = date(2017, 11, 1)
+        stop = date.today().replace(day=1)
         if self.sample:
-            end_date = date(2017, 12, 1)
-        else:
-            end_date = date.today().replace(day=1)
-
-        while current_date < end_date:
-            current_date += timedelta(days=32)
-            current_date.replace(day=1)
+            start = stop
 
-            url = base_url.format(year=current_date.year, month=current_date.month)
-            yield scrapy.Request(
-                url,
-                meta={'kf_filename': hashlib.md5(url.encode('utf-8')).hexdigest() + '.json'},
-                callback=self.parse_list
-            )
+        for d in date_range_by_month(start, stop):
+            yield self.build_request(url.format(d), formatter=components(-2), callback=self.parse_list)
diff --git a/kingfisher_scrapy/spiders/uruguay_historical.py b/kingfisher_scrapy/spiders/uruguay_historical.py
index 1d87df7c..057c2358 100644
--- a/kingfisher_scrapy/spiders/uruguay_historical.py
+++ b/kingfisher_scrapy/spiders/uruguay_historical.py
@@ -1,6 +1,5 @@
-import scrapy
-
 from kingfisher_scrapy.base_spider import ZipSpider
+from kingfisher_scrapy.util import components, date_range_by_year
 
 
 class UruguayHistorical(ZipSpider):
@@ -17,10 +16,12 @@ class UruguayHistorical(ZipSpider):
     }
 
     def start_requests(self):
-        base_url = 'https://www.gub.uy/agencia-compras-contrataciones-estado/sites/agencia-compras-contrataciones' \
-                   '-estado/files/2019-04/OCDS-{}.zip'
-        end_year = 2018
+        start = 2002
+        stop = 2017
         if self.sample:
-            end_year = 2003
-        for year in range(2002, end_year):
-            yield scrapy.Request(base_url.format(year), meta={'kf_filename': 'OCDS-{}.zip'.format(year)})
+            start = stop
+
+        pattern = 'https://www.gub.uy/agencia-compras-contrataciones-estado/sites' \
+                  '/agencia-compras-contrataciones-estado/files/2019-04/OCDS-{}.zip'
+        for year in date_range_by_year(start, stop):
+            yield self.build_request(pattern.format(year), formatter=components(-1))
diff --git a/kingfisher_scrapy/spiders/uruguay_records.py b/kingfisher_scrapy/spiders/uruguay_records.py
index 1e914bd1..90bf601d 100644
--- a/kingfisher_scrapy/spiders/uruguay_records.py
+++ b/kingfisher_scrapy/spiders/uruguay_records.py
@@ -1,9 +1,5 @@
-import hashlib
-
-import scrapy
-
 from kingfisher_scrapy.spiders.uruguay_base import UruguayBase
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import components, handle_error
 
 
 class UruguayRecords(UruguayBase):
@@ -12,12 +8,12 @@ class UruguayRecords(UruguayBase):
 
     @handle_error
     def parse_list(self, response):
-        base_record_url = 'https://www.comprasestatales.gub.uy/ocds/record/{}'
-        root = response.xpath('//item/title/text()').getall()
+        pattern = 'https://www.comprasestatales.gub.uy/ocds/record/{}'
 
+        titles = response.xpath('//item/title/text()').getall()
         if self.sample:
-            root = [root[0]]
+            titles = [titles[0]]
 
-        for id_compra in root:
-            url = base_record_url.format(id_compra.split(',')[0].replace('id_compra:', ''))
-            yield scrapy.Request(url, meta={'kf_filename': hashlib.md5(url.encode('utf-8')).hexdigest() + '.json'})
+        for title in titles:
+            identifier = title.split(',')[0].split(':')[1]
+            yield self.build_request(pattern.format(identifier), formatter=components(-1))
diff --git a/kingfisher_scrapy/spiders/uruguay_releases.py b/kingfisher_scrapy/spiders/uruguay_releases.py
index 832cc027..3e0a6da6 100644
--- a/kingfisher_scrapy/spiders/uruguay_releases.py
+++ b/kingfisher_scrapy/spiders/uruguay_releases.py
@@ -1,9 +1,5 @@
-import hashlib
-
-import scrapy
-
 from kingfisher_scrapy.spiders.uruguay_base import UruguayBase
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import components, handle_error
 
 
 class UruguayReleases(UruguayBase):
@@ -12,10 +8,9 @@ class UruguayReleases(UruguayBase):
 
     @handle_error
     def parse_list(self, response):
-        root = response.xpath('//item/link/text()').getall()
-
+        urls = response.xpath('//item/link/text()').getall()
         if self.sample:
-            root = [root[0]]
+            urls = [urls[0]]
 
-        for url in root:
-            yield scrapy.Request(url, meta={'kf_filename': hashlib.md5(url.encode('utf-8')).hexdigest() + '.json'})
+        for url in urls:
+            yield self.build_request(url, formatter=components(-1))
diff --git a/kingfisher_scrapy/spiders/zambia.py b/kingfisher_scrapy/spiders/zambia.py
index b82185d3..ee2cf877 100644
--- a/kingfisher_scrapy/spiders/zambia.py
+++ b/kingfisher_scrapy/spiders/zambia.py
@@ -3,7 +3,7 @@
 import scrapy
 
 from kingfisher_scrapy.base_spider import ZipSpider
-from kingfisher_scrapy.util import handle_error
+from kingfisher_scrapy.util import components, handle_error
 
 
 class Zambia(ZipSpider):
@@ -19,14 +19,10 @@ def start_requests(self):
 
     @handle_error
     def parse_list(self, response):
-        json_data = json.loads(response.text)
-        files_urls = json_data['packagesPerMonth']
-
+        urls = json.loads(response.text)['packagesPerMonth']
         if self.sample:
-            files_urls = [files_urls[0]]
+            urls = [urls[0]]
 
-        for file_url in files_urls:
-            yield scrapy.Request(
-                file_url,
-                meta={'kf_filename': '%s.json' % file_url[-16:].replace('/', '-')},
-            )
+        for url in urls:
+            # URL looks like https://www.zppa.org.zm/ocds/services/recordpackage/getrecordpackage/2016/7
+            yield self.build_request(url, formatter=components(-2))
diff --git a/kingfisher_scrapy/util.py b/kingfisher_scrapy/util.py
index d425d81c..f20487f5 100644
--- a/kingfisher_scrapy/util.py
+++ b/kingfisher_scrapy/util.py
@@ -1,16 +1,64 @@
 import itertools
 import json
+from datetime import date
 from decimal import Decimal
 from functools import wraps
+from urllib.parse import parse_qs, urlencode, urlsplit
 
 from ijson import ObjectBuilder, utils
 
 
+def components(start, stop=None):
+    """
+    Returns a function that returns the selected non-empty path components, excluding the ``.json`` extension.
+
+    >>> components(-1)('http://example.com/api/planning.json')
+    'planning'
+
+    >>> components(-2, -1)('http://example.com/api/planning/package.json')
+    'planning'
+    """
+    def wrapper(url):
+        value = '-'.join(list(filter(None, urlsplit(url).path.split('/')))[start:stop])
+        if value.endswith('.json'):
+            return value[:-5]
+        return value
+    return wrapper
+
+
+def parameters(*keys):
+    """
+    Returns a function that returns the selected query string parameters.
+
+    >>> parameters('page')('http://example.com/api/packages.json?page=1')
+    'page-1'
+
+    >>> parameters('year', 'page')('http://example.com/api/packages.json?year=2000&page=1')
+    'year-2000-page-1'
+    """
+    def wrapper(url):
+        query = parse_qs(urlsplit(url).query)
+        return '-'.join(s for key in keys for value in query[key] for s in [key, value])
+    return wrapper
+
+
+def join(*functions):
+    """
+    Returns a function that joins the given functions' outputs.
+
+    >>> join(components(-1), parameters('page'))('http://example.com/api/planning.json?page=1')
+    'planning-page-1'
+    """
+    def wrapper(url):
+        return '-'.join(function(url) for function in functions)
+    return wrapper
+
+
 def handle_error(decorated):
     """
     A decorator for spider parse methods.
 
-    Yields a :class:`~kingfisher_scrapy.items.FileError` for non-2xx HTTP status codes.
+    Yields a :class:`~kingfisher_scrapy.items.FileError` for successful HTTP status codes.
     """
     @wraps(decorated)
     def wrapper(self, response):
@@ -21,6 +69,39 @@ def wrapper(self, response):
     return wrapper
 
 
+# https://stackoverflow.com/questions/34898525/generate-list-of-months-between-interval-in-python
+def date_range_by_month(start, stop):
+    """
+    Yields the first day of the month from the ``start`` to the ``stop`` dates, in reverse chronological order.
+    """
+    def number_of_months(d):
+        return 12 * d.year + d.month
+
+    for months in reversed(range(number_of_months(start) - 1, number_of_months(stop))):
+        year, month = divmod(months, 12)
+        yield date(year, month + 1, 1)
+
+
+def date_range_by_year(start, stop):
+    """
+    Returns the year from the ``start`` to the ``stop`` years, in reverse chronological order.
+    """
+    return reversed(range(start, stop + 1))
+
+
+def replace_parameter(url, key, value):
+    """
+    Returns a URL after updating the query string parameter.
+    """
+    parsed = urlsplit(url)
+    query = parse_qs(parsed.query)
+    if value is None:
+        del query[key]
+    else:
+        query[key] = [value]
+    return parsed._replace(query=urlencode(query, doseq=True)).geturl()
+
+
 @utils.coroutine
 def items_basecoro(target, prefix, map_type=None, skip_key=None):
     """
diff --git a/pytest.ini b/pytest.ini
new file mode 100644
index 00000000..df3eb518
--- /dev/null
+++ b/pytest.ini
@@ -0,0 +1,2 @@
+[pytest]
+addopts = --doctest-modules
diff --git a/tests/test_links_spider.py b/tests/test_links_spider.py
index f41a26ee..730cf9e5 100644
--- a/tests/test_links_spider.py
+++ b/tests/test_links_spider.py
@@ -8,12 +8,13 @@
 
 def test_next_link():
     spider = spider_with_crawler(spider_class=LinksSpider)
+    spider.next_page_formatter = lambda url: 'next.json'
 
     request = spider.next_link(response_fixture())
 
     assert type(request) is Request
     assert request.url == 'http://example.com/next'
-    assert request.meta == {'kf_filename': '166715ca8e5f3c1531156d8772b922b7.json'}
+    assert request.meta == {'kf_filename': 'next.json'}
 
 
 def test_parse_404():
@@ -36,6 +37,7 @@ def test_parse_404():
 def test_parse_200():
     spider = spider_with_crawler(spider_class=LinksSpider)
     spider.data_type = 'release_package'
+    spider.next_page_formatter = lambda url: 'next.json'
 
     generator = spider.parse(response_fixture())
     item = next(generator)
@@ -53,7 +55,7 @@ def test_parse_200():
 
     assert type(request) is Request
     assert request.url == 'http://example.com/next'
-    assert request.meta == {'kf_filename': '166715ca8e5f3c1531156d8772b922b7.json'}
+    assert request.meta == {'kf_filename': 'next.json'}
 
     with pytest.raises(StopIteration):
         next(generator)